the valuation of conglomerate companies/67531/metadc...appropriately applied to conglomerates for...
TRANSCRIPT
THE VALUATION OF CONGLOMERATE COMPANIES
APPROVED:
Graduate Committee:
Committee Member
Committee Member
Committee MemSer
Dean of the Softool "of Busine
Dean of th£ Graduate School
THE VALUATION OF CONGLOMERATE COMPANIES
DISSERTATION
Presented to the Graduate Council of the
North Texas State University in Partial
Fulfillment of the Requirements
or the Degree of
DOCTOR OF PHILOSOPHY
By
Winfield P. Betty, B. B. A., M. B. A,
Denton, Texas
May, 1969
Copyright by
Winfield P. Betty
1969
TABLE OF CONTENTS
Page LIST OF TABLES V
LIST OF ILLUSTRATIONS vii
Chapter
I. INTRODUCTION 1
Statement of the Problem Hypotheses Sources of Data Significance of the Investigation Approach
II. SIMPLE RELATIONSHIPS UNDERLYING ACQUISITIONS . . 15
Equations for Growth Trends of Earnings per Share
Effects of Multiple Acquisitions
III. ADDITIONAL BACKGROUND MATERIAL, MOTIVES, AND POSSIBLE EFFICIENCIES UNDERLYING THE GROWTH OF CONGLOMERATE COMPANIES 33
The Attitude of the Supreme Court and Regulatory Agencies toward Conglomerate Companies
Possible Efficiencies Released in Mergers Definition of a Conglomerate Merger Possible Motives of Stockholders and Manage-ment Leading to the Growth of Conglomerate Companies
IV. THEORIES OF VALUE AND THEIR IMPLEMENTATION . . . 52
Theories of Value Investment and Speculative Theories of Value Decision Models Decision-Making Situations Generation of the Relevant States of the World Basic Models Used by Fundamentalists Conflicting Theories of Value The Decision Environment Surrounding
Conglomerate Companies
i n
XV
Page Problems Introduced with the Recognition of Uncertainty
Objectives Underlying the Valuation of Conglomerate Companies
Assumptions about Utility Implied by the Objective to Maximize Expected Present Value
A Theory of Value for Valuing the Stock of Conglomerates
V. BACKGROUND MATERIAL FOR ANALYSIS OF CONGLOMERATE COMPANIES 89
Perfect Expectations Hypotheses Problem Areas in Testing the Hypotheses External Conditions Which Have Facilitated
the Growth of Conglomerate Companies Methods of External Expansion Pooling Versus Purchase Accounting Some Hypothetical Transactions Acquisition Income Acquisition Income of Company M
VI. GROWTH TRENDS UNDER CONDITIONS OF SUSTAINED ACQUISITIONS 126
Difficulties Encountered in the Valuation of Shares in Conglomerates
Sources of Income Value Added Value Added by Company M
VII. RECOGNITION OF PROBLEM AREAS AND TESTING OF THE HYPOTHESES 145
Data Required for Valuing Conglomerates According to the Value Added Criterion
Modifications to the Concept of Value Added Imposed by the Forementioned Problem Areas
Value Added from Stockholders1 Point of View Testing the Hypotheses
VIII. SUMMARY AND CONCLUSIONS REGARDING "VALUE ADDED" BY A SELECTED GROUP OF CONGLOMERATES . . . . 196
Fly po theses
IX. VALUATION ACCORDING TO EXPECTED PRESENT VALUE 208
Valuations in Terms of Expectations Expected Present Value Model
BIBLIOGRAPHY 241
Table
I.
II.
III.
IV.
V.
VI.
VII.
VIII.
IX.
X.
XI.
XII.
XIII.
XIV.
XV.
XVI.
XVII.
. LIST OF TABLES
Page
Effects of Acquisitions on Parent Company . . . 16
Values of Shares Before and After Acquisition. . 17
Synergy Potential 42
Synergy Release 45
Financial Characteristics of Teledyne Incorporated 161
Acquisitions of Publicly Owned Companies Used in Determining "Value Added" 162
Projections Based on Trends Developed Prior
to Acquisition 164
Rates of Change in Real Variables for Teledyne 167
Values for Leverage Variables for Teledyne . . .169
Present Value of Earnings of Teledyne under Conditions of Future Certainty 170
Present Value of Shares of Teledyne under Different Assumed Multiples of Earnings . . .171
Operating and Acquisition Ratios for Teledyne Company 173
Present Value of Stock of Litton under Conditions of Future Certainty 177
Present Value of Shares of Litton under Different Assumed Multiples of Earnings . . .178
Operating and Acquisition Ratios for Litton 179
Operating and Acquisition Ratios for Textron 182
Present Value of Stock of Textron under Conditions of Future Certainty 1.83
v
VI
Table • Page
XVIII. Operating and Acquisition Ratios for Walter Kidde 186
XIX. Present Value of the Common Stock of Walter Kidde under Conditions of Future Uncertainty . . . .187
XX. Operating and Acquisition Ratios for Gulf and Western 189
XXI. Present Value of Future Dividents of Gulf and Western under Conditions of Future Uncertainty 191
XXII. Operating and Acquisition Ratios for Automatic Sprinkler 193
XXIII. Present Value of the Common Stock of Automatic Sprinkler under Conditions of Future Uncertainty 194
XXIV. Comparison of Present Value and Price in the
Year of Valuation 204
XXV. Value Added and Acquisition Strategies 206
XXVI. Value Added and Acquisition Strategies 207
LIST OF ILLUSTRATIONS
Figure Page
1. Possible Trends in Income per Share Resulting from Acquisitions 21
2. Long-run Trends in Earnings per Share if Operating Trends are Negative and Acquisition Income is Positive 29
3. Long-run Trends in Earnings per Share if Operating Trends are Negative and Acquisition Income is Positive with Initial Dilution in Earnings per Share 29
4. Long-run Trends in Earnings per Share if Operating Trends are Positive and Acquisition Income is Negative 30
5. Long-run Trends in Earnings per Share if Operating Trends are Positive and Acquisition Income is Negative with Initial Increase in Earnings per Share 30
6. Merger Spectrum 51
7. Decision Matrix under Conditions of Future
Uncertainty 58
8. General Types of Decision-making Situations. . . 64
9. Specific Types of Decision-making Situations . . 65
10. Types of Decision-making Situations and Methods of Attack . . . . . 66
11. Assumptions Underlying the Major Approaches to Security Valuation 75
12. Matrices Reflecting the Two Major Types of Decision-making Situations 78
13. Indifference Curves Reflecting Constant Expected Present Value; 82
14. Trend in Rate of Return on Investment for Company M 121
VI1
Vlll
Figure Page
15. Trend in Total Investment for Company M . . . .121
16. Trend in Total Revenue for Company M 122
17. Trend in Fixed Charges for Company M 122
18. Trend in Net Contribution to Equity for Company M 123
19. Trend in Net Contribution to Equity Minus
Preferred Income for Company M 123
20. Trend in Earnings per Share for Company M . . .124
21. Trends Reflecting Positive Acquisition Income from Single Acquisitions 12 3
22. Long-run Trends Resulting from Sustained Acquisitions Which Add Immediately to Earnings per Share 129
23. Long-run Trend in Earnings per Share When Acquisition Income is Positive and gs is Greater Than gp 131
24. Gap Between Operating Trend and Total Income Trend .133
25. Gap Between Operating Trend and Total Income Trend 136
26. Divergence Between Past Trend and Operating Trend .140
27. Relevant Time Periods in the Determination of "Value Added" 153
28. Valuation Matrix under Conditions of Future
Uncertainty 215
29. Initial Range of Expectations 223
30. Bisection of the Initial Range of Expectations 224
31. Range of Expectations for Rate of Return on Investment for All Nonfinancial Corporations . 225
IX
Figure Page
32. Expected Growth Rate in Rate of Return on
Investment (gRinf) 226
33. Expected Growth of Investment (glnf) 226
34. Range of Possible Expectations for Ia in Millions of Dollars 232
35. Range of Possible Expectations for Growth in Total Investment of Future Acquisitions after Acquisition (gla) 232
36. Range of Possible Expectations for Rate of Growth in Return on Investment of Acquisitions After Acquisition (gRia) 232
37. Range of Possible Expectations for Rate of Return on Investment of Acquisitions in the Year of Acquisition (Ria) 233
33, Yields on Preferred Stock (Expected) 237
39. Rate Paid per Share to Minority Interest (Expected) 238
CHAPTER I
INTRODUCTION
Business combinations (generically referred to as
mergers) are not a new phenomenon in the United States.
Since 1900 there have been three distinct movements, and each
of these movements has had its unique motivating forces.
The first wave of mergers occurred near the turn of the
century and was apparently motivated by desires of business-
men to limit competition from similar firms. Since the
motivating force is manifested in the form which the merger
assumes, it is not surprising that the dominant form of this
period was the horizontcil one, in which similar firms were
combined.
The second major merger movement occurred during the
period from 1917 to 1929. Undoubtedly a certain amount of
the merger activity of this period was speculatively moti-
vated. The emergence of the massive public utility holding
companies during this period is probably adequate evidence
of such motivation. However, other mergers occurred which
were the results of attempts of businessmen to assure them-
selves of dependable sources of supplies, or alternatively,
to attain efficiencies in distribution processes. When
motivated by these latter factors, the combinations which
resulted were of the vertical or circular variety.
In the third merger movement (1948 to the present) the
motivating factors and realities are not quite as clear as
in the former cases because as complete a historical per-
spective is not yet available. Several aspects of this most
recent movement, however, are becoming apparent. First,
while some of the activity is of the horizontal, vertical,
and circular variety, there is an increasing tendency for
firms operating in different markets to be joined together
into conglomerate-type companies. •*• Furthermore, this mode of
expansion is gaining momentum in the 1960's, as the following
quotation indicates.
Some new statistics on the galloping merger movement sent echoes through Washington this week. Even antitrusters were startled by the totals in the Federal Trade Commission's annual report on corporate mergers. Its highlights:
During 1967, mergers among U.S. companies humped 37% to an all-time high of 2,384.
Acquisitions of large companies with assets of $.10 million or more dominated. There were 155 of these large acquisitions last year, compared to 101 in 1966. Of the 155, conglomerates accounted for 128, up from 75 in 1966.
Companies with assets over $100 million acquired 62% of the large companies involved in mergers; these companies acquired more bhan $6.3 billion, or 79% of the assets.2
-*-See page 34 for a definition of conglomerate merger,
^"Measuring the Surge of Mergers," Business Week, No. 2013 (March 23, 1968), p. 69. " ~
It is possible that conglomerates are being created
because businessmen are unwilling to make heavy commitments
of resources to basic research unless they are somehow
guaranteed a degree of size and stability of earnings not
attained by a firm operating in a single market. Another of
numerous possibilities is that in a technically complicated
environment such as today's, the level of managerial skill
demanded has made management a scarce commodity. Under these
conditions it is possible that diverse firms joined together
under a single management could operate more efficiently than
the components alone. In any event, the precise forces
underlying the current movement are uncertain, and the fore-
mentioned reasons are mentioned only because they represent
plausible possibilities.
Statement of the Problem
A basic question remains unanswered—Is the current
merger trend resulting in improved income potential of the
combined resources, or is it merely the manifestation of
promoters' desires to make a quick profit? This question and
others remain largely unanswered because traditional methods
of analysis are not applicable to conglomerate companies.
This lack of applicability has left a gap in the ability of
investors to identify conglomerates of investment-grade
quality.
4
Valuation techniques traditionally used in the analysis
of companies operating in a single industry cannot be
appropriately applied to conglomerates for the following
reasons:
(1) The operating data for conglomerates are usually
furnished on a consolidated basis only. (If operating results
were provided for each subsidiary, then each might be valued
independently.)
(2) Diversification of earnings is often cited as one
of the advantages of conglomerates as an investment; however,
no tools are available for placing a value on this charac-
teristic (£.e., the quality and stability of total earnings
from diversified sources).
(3) Current earnings and projected changes in earning
levels are the two main independent variables used in
traditional methods of valuation. For several reasons
difficulties are encountered when attempting to project the
earnings of conglomerates. First, most conglomerates have
emerged in recent years and have not been tested by adverse
business conditions. Second, the earnings reported by con-
glomerates are derived not only from operating sources but
also from acquisitions. For example, LTV, in the period
1957-1966, increased its total earnings by 3,900 percent"^—
a growth rate that would be very difficult to attain by'
3 "James C. Tanner, "Ling's Empire," Wall Street Journal,
CLXX (August IB, 1967), p. 1. ~
internal means alone. Furthermore, this mode of growth can
be transferred to parent company shareholders if subsidiary
companies are obtained under favorable financial terms.
Accordingly, on an overall basis or on a per-share basis,
earnings of conglomerate companies can be greatly influenced
by the mere act of acquiring resources from external sources.
It is possible for a conglomerate company to augment
its income from operations by making acquisitions. That is,
changes can occur in reported income and profitability because
of both operating and acquisition activities. Furthermore,
a company relying solely on acquisitions can create an upward
trend in profitability as measured by rate of return on
investment. Not only is it possible for increases in rates
of return on investment to be passed on to common stock-
holders, but also, such increases can be greatly magnified
in terms of earnings per share. Such impressions of growth
can occur when, in reality, the productivity of resources
employed is declining. The source of the deception lies in
the calculation of growth rates relative to the past time
period and not relative to the past rate of return on the
resources employed (i^e., traditional calculations are not
geared for companies which acquire most of their productive
resources second hand).
Hypotheses
To facilitate the investigation of the sources of growth
which are available to conglomerate companies and to draw
some limited conclusions with regard to which are the major
sources, several hypotheses are presented and tested. These
hypotheses are as follows:
(1) Conglomerate companies listed on the New York and
American Stock Exchanges acquiring the highest proportion of
publicly owned subsidiaries in the period 1961-1967 have
added negative value to them—where positive or negative
value added is measured relative to the weighted average of
the rates of return on book values of the resources employed
by the subsidiary companies prior to their acquisition.
Accordingly, negative value added is defined as the extent
by which the projected rate of return on investment exceeds
the rate actually experienced.
(2) The value added by conglomerate companies during
the period 1961-1967 has tended to decrease from year to year
as size and diversification have increased—where diversifi-
cation is measured by the number of different industries in
which resources are employed, and size is measured in terms
of the book value of consolidated assets. (The decrease in
value added can be from negative values to more negative
values or from positive values to less positive values.)
(3) Eased on a projection of past acqui.sit.ion policies,
past weighted average growth trends of parent -and subsidiary
companies, and the same percentage of value added by parent
company organizations, the current market prices of the
common stocks of conglomerate companies are too high if an
8 percent rate of return is demanded.
(4) If the conglomerate companies surveyed are grouped
into (1) those adding the most value (relative to the other
companies) and (2) those adding the least value, then the
companies in each group will tend to have similar acquisition
strategies. Acquisition strategies for this purpose are
defined in terms of the relative rates of return and growth
rates of subsidiaries and parent companies at the time of
acquisition.
The testing of the hypotheses is intended more as a
vehicle for illustrating the possible avenues of growth
available to conglomerate companies than as a means of
drawing conclusions of a general nature. The results derived
in testing the hypotheses are of limited validity for at
least two reasons. First, the companies used in testing the
hypotheses are not selected in a statistically random manner,
but rather, consideration is limited to a few of the com-
panies which have acquired a relatively large number of
publicly owned companies. Also, in testing for "value added"
attention is directed only to the publicly owned companies
acquired by a particular conglomerate company and not to all
of its acquisitions.
8
Sources of Data
The data used for testing the hypotheses are derived
from several sources. The major source is Moody's Industrial
Manuals; however, reference is also made to the 10-K reports
filed with the Securities and Exchange Commission, to annual
reports of the specific companies, and to numerous profes-
sional journals. Also, Moody's Handbook of Common Stocks
anc^ Standard and Poor' s Security Owner' s Stock Guide are used
to determine price-earnings multiples and prices of shares
which prevailed in different years.
Significance of the Investigation
The rapid growth rate in the size of individual con-
glomerate companies coupled with the absence of adequate
tools for their analysis has created problems of increasing
dimensions for investors and public-policy makers alike.
That is, as the conglomerate form of organization has con-
tinued to grow, the problems of analysis which they present
have also been compounded.
Almost all of the investigations conducted to date by
academicians and public-policy makers have proceeded along
traditional lines. Accordingly, such investigations have
not given recognition to the unique characteristics of
conglomerate companies. The major significance of this
investigation is that it does recognize the uniqueness of
conglomerate companies, and the conclusions derived, if they
are valid, can be helpful in making future decision making
more meaningful.
Approach
The investigation is structured on the basis of a set
of relationships which are derived and expressed in succeeding
chapters. The general format of the investigation is to
develop basic concepts and relationships in the initial
chapters, and in subsequent chapters to make added refine-
ments which will clarify all of the sources of growth
available to conglomerate companies. Finally, after all of
the relevant factors influencing the growth potential of
conglomerate companies have been recognized, a method for
valuing the shares in conglomerate companies according to the
present value criterion is illustrated.
The approach used in the investigation can perhaps be
best summarized by a brief examination of the basic relation-
ships which are derived and expressed in the form of
mathematical equations. All of the equations and their
related implications are discussed briefly in the subsequent
paragraphs. The equations are as follows:
(D Ya = Tesd+gs^'Z + E p U + gp)b'Y
Y + Z
C2) Tt = EP ( 1 * gP* b'Y + Te (l+gg)
k*Z
Y + tz k=o ~Y~rrr"
10
(3) E g t = (Tes ~ Yt*)Z
Y + tZ
C4) Yj. = (i-T) idp) (RId) + ZI d a ) CRia)-(Rd) (D)J-(V (p) c f f a = Q
a=t (5) Rt = k ' V <
RIp» +^_Ia(l+gIa>t"a-Rla(1+9Rla) t~~a
In +llla(1+gIa)t"a
"P a^b a
(6) Yt = (R
0) ( V + t^tx) (^t) (IQ) (t-t0)-(Rp) (PQ)
n=rn (7) E p v = (PrJ (PVn)
n=l rn
Equation (1) is used to define the effects of various
acquisition strategies on earnings per share. In this
equation Ya represents future levels of earnings per share
which result from trading Y shares of a parent company for
the shares of stock of a desired acquisition candidate. T e g
refers to the earnings per share of the acquired company
after the shares of the parent company have been traded for
its shares, Z refers to the number of shares of stock out-
standing in the parent company prior to acquisition of the
additional company, gs is the growth rate of the parent
company's earnings per share, Ep is the earnings per share
of the initial parent company, and t is time, measured from
f1nQ nn i nf r\ f ^ rrrn i o i S -r~\y*
11
As explained more fully in Chapter II, if T e s is greater
than Ep, there will be an immediate increase in earnings per
share from acquisitions; also, depending upon whether or not
g is greater than gp, the combined operating trend will be
improved or diluted following acquisition. Based upon the
relative values for T e s, Ep, gg, and gp—six possible
acquisition strategies are defined, and resultant values for
Ya are presented graphically.
Eguation (1) is used to define the results of single
acquisitions, whereas equation (2) represents the effects on
earnings per share (Yt) of sustained acquisitions when
relative values for Tes, Ep, gs, and gp are assumed to remain
constant. Using equation (2), various possible growth
patterns are illustrated graphically. Of these patterns,
the most significant is the one illustrating that growth
through acquisitions alone cannot be sustained indefinitely.
That is, when T e s is greater than Ep but gs and gp are
negative, only a temporary positive growth trend can be
established.
Equation (3) is used to show that even when both
acquisition income and operating trend are positive (i.e.,
when T e g is greater than Ep and gp and gs are both positive),
the growth pattern in the early periods of acquisition cannot
be maintained unless the size, rate, or terms of acquisitions
are continually improved. Accordingly, regardless of the
values for Tes, Ep, gs, and gp—the growth trends developed
12
in the past by conglomerate companies are not necessarily a
meaningful basis for making future projections.
Equation (4) is used to introduce additional variables
which influence earnings per share of conglomerate companies.
Using this equation, attention is directed to rates of return
on investment and total investment; accordingly, attention
is directed away from a "per-share" orientation. In
equation (4) T refers to the income tax rate, Ip is total
investment of the initial parent company, Rip is rate of
return on investment of the initial parent company, Ia is
the average size of investment of future acquisitions, Rja is
the rate of return on investment of future acquisitions,
is the average rate paid to debt in the future, D is the
total amount of debt employed, Rp is the rate paid to pre-
ferred stock holders, P is the total amount of preferred
outstanding, and S refers to the total number of common
shares outstanding.
Equation (4) is used to illustrate how a parent company
with a declining rate of return on investment could show a
positive trend in rate of return on investment .in spite of
its continued acquisition of companies whose rates of return
on investment were also declining. Furthermore, it is shown
how the false impression of growth can be transferred to
earnings per share and possibly magnified by changes in R^,
D, Rp, P, and S.
13
Equation (5) is introduced to test whether or not value
was added by a selected group of conglomerates to the
publicly owned companies which they acquired. Value added
was considered positive anytime that K in this equation was
greater than one, and value added, negative, anytime that the
value for K was less than one.
Equation (6) is used to project future levels of
earnings per share under the assumption that future changes
in rate of return on investment after taxes will remain the
same as in the past. In this equation R -x refers to rate of
return on investment after taxes, It refers to total invest-
ment, and A^t is the rate of change in return on investment K it
per unit of investment. Also, t refers to time, and A*t At"
represents the rate of change in investment per unit of time.
All of the variables with subscripts of "o" are values which
exist in the year that valuation occurs.
Equation (6) is used to estimate the investment value
of the shares of a selected group of conglomerates under the
assumption that such value is derived from the sale of the
stock at the current price-earnings multiple after a ten-year
holding period. Using these results, an attempt is made to
determine whether or not such stocks are currently over-
valued.
Finally, in the last chapter, equation (7) is used to
show how the shares of stock in conglomerate companies can be
.14
valued in accordance with the "resent-value" criterion.
In this equation EpV refers to expected present value, m
represents the number of sets of future conditions which are
considered relevant possibilities, and P r is the present
value of the stock under the assumption that conditions "n"
actually exist.
CHAPTER II
SIMPLE RELATIONSHIPS UNDERLYING ACQUISITIONS
A simple example should clarify how a conglomerate
company, through its acquisition strategy, is able to give
the impression of being a growth company when it is actxaally
declining in terms of growth in earnings of the resources
employed. Assume, for example, that a parent company has
30,000 shares of common stock outstanding and earns income
at the rate of $10 per share per year with no growth. Assume
further that the company acquires one additional subsidiary
in each of four consecutive time periods, and that the sub-
sidiaries acquired have 10,000 shares outstanding. Assume
also that each subsidiary earns $20 per share in the year it
is acquired, and that each one has a negative growth trend
in earnings per share of $2 per year. Finally, assume that
the investment community views the parent company's earnings
record as indicative of future growth, and that its stock is
selling at a high enough price so that it is able to acquire
the shares of the four companies by swapping shares of stock
on a one-for-one basis. Under these conditions, the. results
found in Table I were obtained.
The average increase in earnings per share attributed
to the parent company's stock is roughly 9 per cent per
15
16
TABLE I
EFFECTS OF ACQUISITIONS ON PARENT COMPANY
Period 0
Period 1
Period 2
Period 3
Period 4
Earnings from subsidiaries
$200,000 $200,000 180,000
$200,000 180,000 160,000
$200,000 180,000 160,000 140,000
Earnings from parent company
$300,000 300,000 300,000 300,000 300,000
Total 300,000 500,000 680,000 860,000 1000,000
Shares outstanding
30,000 40,000 50,000 60,000 70,000
Earnings per share
$10.00 $12.50 $13.60 $14.34 $14.30
period, as shown on the bottom line of the above table. This
growth trend appeared in spite of the fact that the real
productivity of the resources employed was declining. Not
only are the earnings reported misleading, but also, the
valuation placed on the stocks will likely be out of line
relative to real growth potential.
Assuming that the market value of the subsidiary
company's shares in the year of acquisition is eight times
current earnings, then the parent company multiple could be
assumed to be around sixteen, since acquisition occurred on
a two-for-one basis in terms of current earnings. If these
17
assumptions are true, the values of the shares before acqui-
sition and afterwards are given by the following:
TABLE II
VALUES OF SHARES BEFORE AND AFTER ACQUISITION
Period 0
Period 1
Period 2
Period 3
Period 4
Parent before acquisition
$160.00 $200.00 $217.60 $229.49
Subsidiary before acquisition
160.00 160.00 160.00 160.00
Parent after acquisition
200.00 217.60 229.49 228.80
Subsidiary after acquisition
Same as parent
Same as parent
Same as parent
Same as parent
The implication of the foregoing discussion is that any
investment in the shares of conglomerate companies should be
made only after a clear distinction has been made between
operating earnings and earnings derived from acquisition.
This separation is necessary if one is to place meaningful
values on such shares. The next section of this thesis
outlines some of the basic relationships needed for analysis
of conglomerate companies.
18
Relationships Underlying Single Acquisitions
Definition of terms:
E . . . Current earnings per share of parent company P
Es . . . Current earnings per share of company to be acquired
Tes. . . Earnings per share of company to be acquired after shares of the parent company have been traded for its shares
gp . . . Growth rate of parent company
gs . . . Growth rate of acquired company
Y . . . The number of shares of the parent company before acquisition
Z . . . The number of shares traded of the parent company for those of the acquired company
Eg . . . Earnings per share added to the shares of the parent company because of the acquisition
Ys . . . Projected earnings per share of the acquired company considered separately after acquisition
Yp . . . Projected earnings per share of the parent company considered separately after acquisition
Ya . . . Actual earnings per share projected for the parent company and the acquired company combined
The symbols above are used to describe some of the basic
relationships inherent in any combination. The situation
described is a simple acquisition in which a certain number
of common shares of the parent company are traded for those
of the company being acquired. While the conclusions apply
to the specific situation described, the same general results
19
can be obtained for more complex arrangements by making
slight modifications in the basic models.
Income Effect
In any acquisition involving a simple trade of common
stock, so long as T e s is greater than Ep, there will be an
immediate increase in the earnings per share of the parent
company. On the other hand, if Ep is greater than Tes/ there
will be an immediate dilution in earnings per share of the
parent company. The use of the income effect in a positive
manner by the parent company is easiest when the parent
company has a higher price-earnings ratio than the company
being acquired. Under these conditions, the parent company
is normally able to trade fewer shares relative to earnings
than it receives.
In addition to data regarding the current earnings per
share of both parent and subsidiary company, as well as
information about the number of shares traded between the
two companies, a more complete description of the combination
requires a specification of the expected growth rates- of both
companies.
In terms of relative growth and relative earnings per
share, there are six possible configurations which a combi-
nation might assume. These configurations are
(l)a Positive Immediate Income Effect (1) 9s ~ 9p
^ (l)^ Negative Immediate Income Effect
20
(2) g s > g p
(2) _, Positive Immediate Incorae Effect CI
(2)^ Negative Immediate Income Effect
(3) g s < g P < (3) Positive Immediate Income Effect
(3)k Negative Immediate Income Effect
Graphically, these six situations are described in Figure 1
on pages 21 and 22.
Eg, the income effect, is added to or subtracted from
the current level of earnings of the parent company to arrive
at the adjusted earnings per share after the combination.
The magnitude of the income effect can be calculated using
the following formula:
Current combined Total earnings of earnings of parent parent and subsidiary and subsidiary's common - calculated at earnings stock in total per share rate of parent
company prior to acquisition Eg =
Total number of shares of common stock after combination
(Tes) (Z) + (Y) (Ep) ~ (Y + Z) (Ep)
(Y + Z)
Simplifying
Z (Tes - E p) Eg =
Y + z For income effect to be positive, Te must be greater than'Ep.
Equations for Growth Trends of Earnings PerShare
The general equations for the future earnings per share
of the subsidiary, the parent company, and their combined
21
Positive Income (Eg) Effect
Earnings per Share
Time
Negative Income Effect
(Eg)
Earnings Share
JL irne
U)b
Positive Income Effect
(Eg)
Earnings per Share (2)
Time
Fig. 1—Possible trends in income per share resulting from acquisitions.
Negative Income Effect
Earnings per Share
22
( 2 ) .
/ / Ua)
(Eg) I
Time
Positive Income (E ) Effect g
Earnings per Share Ya)
( 3 )
Txme
Negative Income (E ) Effect
Earnings per Share
0 Time
Fig. l-~Continued
(3) k
23
earnings per share are given in the following three equations
respectively.
(1) Ys = T e s (1 + gs) t
(2) Yp = Ep (1 + gp) t
(3) Y a = (1 +.gS)t Z + Ep (1 + gp)t Y
Y + Z
To check equation (3) we know that when t equals zero, the
correct value for Ya is Ep plus or minus Eg. Accordingly,
when t equals zero,
Y a = (Te3) (Z) + (
Ep) (*)
Y + Z
Simultaneously adding and subtracting (Z) (Ep) in the
numerator of the above equation causes it to assume the
following form:
Y a = (Tes) (Z) - (Z) (Ep) + (Z) (Ep) + (Ep) (Y)
Y + Z
Writing the equation in two parts and simplifying,
Y a = Z (Tes - EP> + Ep (z + Y)
_ — y " T 1 '
The equation then simplifies to
Y a = Eg + Ep (when t equals zero)
24
Time until Break Even
Under certain circumstances (such as situation 2^,
described in the diagrams previously presented) the parent
company will accept a negative income effect with the expec-
tation that the higher growth rate of the acquired company's
resources will more than justify the initial dilution in
earnings per share, or possibly (as in situation 3a) a
company will be attracted by the favorable income effect and
assume that the initial increase in earnings per share will
more than justify the loss in earnings per share at some
future date. In any case, the time until the income effect
is offset by growth is an important parameter. It is calcu-
lated in the following manner:
Break even occurs where Y^ = Yp or, alternatively, v/here log Ys = log Yp. Since
Ys = T e s (1 + gs)fc
and
Yp = Ep (1 + 9p)^
also
log Ys = t log (1 + gs) + log Tes
and
Er
therefore, at breakeven
t log (1 + gs) + log T e s = t log (1 + gp) + log Ep
solving for t
log Yp = t log (1 + gp) + log
25
t = log Ep - log T e g
log (1 + gs) - log (1 + gp)
Note that in the preceding equation for t, the only
independent variable subject to control is T e s, the trans-,
posed earnings per share of the acquired company. Since
W Tes = % (Es)
Where
W . . . refers to the number of common shares out-standing in the subsidiary company before acquisition
Z . . . as before, refers to the number of shares of the parent company traded for the shares of the subsidiary
Eg. . . refers to current earnings per share of the subsidiary before acquisition and measured in terms of its own shares
Accordingly, Tes can be controlled by adjusting the ratio of
Z to W. By doing this, the time to break even can also be
shortened or extended. In fact, the appropriate exchange
ratio could be established on this basis if one arbitrarily
adopted some minimum time until break even.
In a broader context., it might be possible to say that
gs and Es are variables subject to control. This is true in
the sense that one might not agree to acquire companies
unless their current earnings per share and their growth
rates were above some given level. If this were the case,
then one wishing to minimize the time until break even (2^)
would try to maximize the ratio of g s to g p and minimize the
26
ratio of Ep to Tes. On the other hand, if one wished to
mazimize the length of time until break even, then an attempt
would be made to minimize the ratio of gs to gp and to
maximize the ratio of Ep to Tes.
Effects of Multiple Acquisitions
All of the preceding equations express relationships
which apply to a single acquisition. The growth trend of
earnings of a conglomerate company with only one subsidiary
was denoted by the following equation:
Y = lT e s (1 + 9s)t Z] + [Ep (1 + g p)t Y]
[Y] + [Z]
If there were more than one acquisition, the total earnings
could be represented by an equation of the same form but with
an additional term in both numerator and denominator for each
of the additional subsidiaries.
The growth trend of a conglomerate is actually the
weighted average of the growth trends of the individual
subsidiaries. The reason that a conglomerate company can
sometimes give the impression of growth when growth does not
exist is that sometimes it is able to acquire subsidiaries
which contribute more to current earnings than the overall
growth rate takes away. Such a policy of acquisition by a
parent company becomes progressively more difficult to pursue
in a successful manner because the weighted-average growth
rate continues to become more and more negative as companies
27
are added. Since the rate of return in any time period is
the weighted average of all components, larger and larger
companies must be acquired in order to offset the effects of
the real decline in productivity. Beyond some point the
company is unable to acquire companies of the magnitude
required.
Earnings Patterns under Conditions of Multiple Acquisitions
Assume that a parent company has current earnings per
share of Ep and a growth rate gp. Also let this company
acquire each year, on the average, a subsidiary whose trans-
posed earnings per share are Tes and a growth rate in earnings
per share of gs.
Yp = Ep <1 + gp)fc Y
and
Ys = Tes (1 ^ 9s)^ 2 #
Letting Yt denote total income per share, then
When t = 0 Yt = YP = EP ' Y
When t = 1
When t = 2
Y1 = Ep (1 + gp)l • Y + T e s • Z
Y + Z
E (1+g p)2 • Y+Tes * Z+Tesd+gs)1 * Z
Y O = —A~——£-Y + 2Z
28
When t = 3
eP ( 1 +V 3' Y + 'es'Z + 'esd+Ss)1"2 + T e s(l+g s)2'Z
Y 3 Y + 3Z
or, in general terms:
k = t - 1 Ep(l+gp)t-Y 'SZ Tes (l+gs)
k*Z = __L _ + k = o
Y + tZ Y + tZ
The relationship expressed in the above equation, when
graphed as a function of time, will take on different shapes,
depending upon the values assigned to Ep, T e s, gp, and gs.
Generally, if gp and gs are both positive and T e s is greater
than Ep, the curve will have a continuously positive slope as
t increases. If, on the other hand, T e s is greater than Ep
but gs and g^ are negative, the curve will have a positive
slope only so long as acquisition income is greater than
the decrease in income per share from the negative growth
rates. Depending upon precise values for the variables,
the curves will appear similar to one of the following if
acquisition income is positive but overall growth rates are
negative.
The curve could have the shape shown in Figure 3 because
of the possibility that the parent company might have a very
high declining growth rate (percentage-wise) and, therefore,
higher dollar losses in income in the earlier years than in
29
(1) Earnings per Share
E. P
0 time
Fig. 2—Long-run trends in earnings per share if operating trends are negative and acquisition income is positive.
Earnings per Share
( 2 )
Ev
0 time
Fig. 3--Long-run trends in earnings per share if operating trends are negative and acquisition income is positive with initial dilution in earnings per share.
subsequent years. Under these conditions the initial losses
.in income by the parent company could be greater than the
gains in acquisition income, but, as these losses decreased
in dollar terms, it would be possible for acquisition income
to cause total income per share to increase for a time period.
30
Finally, as less and less income is derived (percentage-wise)
through acquisition, the overall declining growth rate will
once again cause earnings per share to decline absolutely.
Under conditions in which the growth rate of parent
company and subsidiaries are both positive, but acquisition
.income is negative, curves similar to the ones above would
be appropriate except the curves would be inverted. That is,
when T e s is less than Ep but gp and gs are positive, the
following growth patterns are possible.
Earnings per Share
E_
time
Fig. 4—Long-run trends in earnings per share if operating trends are positive and acquisition income is negative.
Earnings per Share
E P
time
Fig. 5~-Long~run trends in earnings per share if oper-ating trends are positive and acquisition income is negative with initial increase in earnings per share.
31
Figure 4 depicts a situation in which acquisition income
dilutes earnings per share until such time that the growth
in operating income per share exceeds the dilution. In
Figure 5, growth in operating income per share is initially
a larger absolute amount than the dilution in earnings per
share experienced through acquisitions. Accordingly,
earnings per share increase until dilution from acquisitions
exceed the growth in operating earnings per share. There-
after, the pattern is the same as that shown in Figure 4.
It can be shown that as long as gs and gp are negative,
the value for the expression
k = t--l E (l+gp)
t-Y Tes(1+9s>k*z v-1-1 yp
= + Y + tZ Y + tZ
Income from Income from Parent Subsidiaries
converges as t increases. That is, it can be shown that the
eventual downward slope of the line depicted in Figures 2
and 3 is correct. Accordingly, acquisition income alone
cannot be used to sustain growth indefinitely under such
conditions.
It is fairly obvious that the contribution of the parent
company to total earnings approaches zero as time increases
and so long as gp is assumed to be negative. However, it
is not as obvious in the case of the income provided by the
subsidiary companies. The denominator of the fraction
32
expressing the income contribution of the subsidiaries
becomes progressively larger as t increases; accordingly, if
the numerator converges, the entire expression converges.
The series k = t-1
^ 1 1 Tes(1+gS)k"z
is a geometric series of the form n
where
a + ar + ar^ + ar^ arn"l
a = Tes•Z and r = (l+gs)
Since this is a geometric series, the absolute value of r
determines whether or not the series converges. If the
absolute value of r is greater than one, the series diverges;
if it is less than one, the series converges. Since it was
assumed that gs is negative, it follows that the value for r
is negative, the series converges, and the value of the
total function decreases as time increases.
CHAPTER III
ADDITIONAL BACKGROUND MATERIAL, MOTIVES, AND
POSSIBLE EFFICIENCIES UNDERLYING THE
GROWTH OF CONGLOMERATE COMPANIES
Perhaps the tone of this paper has been unduly harsh
regarding the motives underlying the rapid growth of conglom-
erate companies—implying that the attainment of acquisition
is the only possible motive. In fact, neither the motives
nor the circumstances surrounding conglomerates are simple,
nor are they necessarily the same in all cases. The study
and analysis of conglomerates is complicated by the following
factors:
1. No unanimity of purpose on the part of those forming
conglomerate companies
2. No generally accepted definition of what constitutes
a conglomerate company
3. The preoccupation of government agencies and the
courts with maintaining competition
4. The possibility that efficiencies in operation are
derived through conglomerate mergers
5. The lumping together of operating and acquisition
income
6. The absence of established earnings trends for
33
34
conglomerate companies which include periods of adverse
business conditions
7. The availability of financial data for conglomerates
only on a consolidated basis
8. Lack of uniformity of accounting methods between
companies
9. Lack of comparability of circumstances before and
after merger
10. From the viewpoint of an investor, the lack of a
theory of value for placing a value on the shares of conglom-
erate companies in light of the forementioned difficulties
and in light of the uniqueness of conglomerate companies.
The obstacles confronting anyone who purports to draw
generalizations about conglomerate companies are exceeded in
magnitude only by the need of such generalizations. While
no study can solve all the forementioned difficulties, it is
possible to circumvent enough of them so that the investor
can base his decisions on the "present value" criterion. At
least this is the main objective of this paper--to provide
a way for investors to value conglomerate companies. The
study does not claim to solve all the issues from an economic
point of view.
In remaining chapters, attention is devoted to all ten
of the problem areas listed above. In the following section
of this chapter, there is a discussion of the attitude of
the courts and public-policy maker, a discussion of possible
35
efficiencies underlying the conglomerate-type mergers, a
discussion of some of the more obvious motives of those
responsible for the formation of conglomerate companies, and,
finally, a definition of conglomerate mergers. The other
issues are treated in subsequent chapters.
The Attitude of the Supreme Court and Regulatory Agencies toward Conglomerate Companies
Because the long-run survival of conglomerate companies
rests on their legal sanction, it is impossible, or at least
illogical, to ignore the attitude of the courts and public-
policy makers. Thus, the following is a brief explanation
of the apparent attitude of these parties.
The growth of conglomerate mergers has created some
unique problems in the formulation of public policy. The
Federal Trade Commission and the Antitrust Division of the
Justice Department share the responsibility for approving
and regulating mergers; however, in contested cases the
Supreme Court is the final authority. Accordingly, these
agencies tend to look toward the courts for clarification.
Traditionally, the courts have applied specific criteria
to determine whether or not a substantial reduction in
competition has resulted from a particular merger (the
criteria relate to the number of firxas and share of market
affected). Since past considerations have been largely
limited to traditional forms of mergers, existing guidelines
are rather awkward when applied to conglomerate mergers. In
the spring of 1968 Donald F. Turner, Assistant Attorney
General who was in charge of the Antitrust Division,
questioned his own position in moving against conglomerates,
stating that these laws are "no panacea. The Federal Trade
Commission chairman has expressed a similar reluctance to go
after conglomerates, stating that he needs the majority of
the support of the four other commissioners.^
The courts have been placed in a rather strained
position since they must apply laws and guidelines not really
designed for application to conglomerate-type companies.
(With conglomerate mergers the number of competitors remains
the same, and no actual change in marketing structure occurs,
initially.) The Supreme Court, in its decision to require
Procter and Gamble to divest itself of Clorox, extended the
concept of "relevant market" to include potential as well as
actual competition.3 That is, the court held that Procter
and Gamble was a potential entrant into the bleach market
because it was a natural extension of their existing product
lines. The court called this a "product extensions" merger.^
1,1Measuring the Surge of Mergers," Business Week, No. 2013 (March 23, 1968), p. 69.
^Ibid.
3James M. and D. Jeanne Patterson, "Conglomerates: The Legal Issues," Business Horizons, XI (February, 1968), p. 42. '
4Ibid., p. 43.
37
In a Circuit Court ruling in the General Foods-S.O.S.
decision, Judge Austin L. Staley extended the concept of
"relevant competition" even further when he stated that
The commission could reasonably conclude that the potential competition was adversely affected by General Foods' entrance into the market through the acquisition of S.O.S. because the entry of such a large, well financed, aggressive competitor would necessarily hamper whatever effect potential competition had in the pre-merger market.5
Apparently, the guidelines being followed by the courts
regarding conglomerate mergers relate to the nearness of the
merger to either a horizontal or vertical merger—i_.e. , a
product extension merger-—or else to the size and power of
the company involved. The position of the Justice Department
was somewhat clarified on May 31, 1968, with the publication
of a set of merger guidelines which listed the ones likely
to be challenged. Specifically, the guidelines state that
the following types of mergers will continue to be chal-
lenged:
Those so-called horizontal mergers involving direct competitors, where in highly concentrated markets the acquiring company and the acquired concern each account for as little as 4% of their market.
"Vertical" mergers involving suppliers and their customers such as a shoe manufacturer and shoe retailers, where the supplier accounts for 10% or more of total market sales and the customer accounts for 6% or more of total market purchases.
And "conglomerate" mergers, defined as those that are neither horizontal or vertical, where a
^Ibid., p. 42.
38
very large company that might have entered a new business through internal expansion instead enters by acquiring an important concern already in the business; where the merger creates dangers of re-ciprocal buying, and where the acquiring company's promotional resources are so great as to enhance an acquired company's existing dominance in its
field.6
The guidelines just presented might imply that the
Justice Department has moved against conglomerate mergers as
well as other forms; however, this is not the case, as the
following indicates: The Clayton Act since 1950, has been used
by the Justice Department and the FTC to stop almost all proposed horizontal mergers, and the stringent guidelines concerning such mergers are taken directly from these cases. The act has been similarly used against vertical mergers and, as a result, very few corporate merger proposals nowadays are of either type. The great majority of mergers are the conglomerate type. The FTC reported recently that 83% of all "large" mergers in 1967 were conglomerates.7
The guidelines which have been promulgated and the recent
court decisions suggest that the legality of a particular
conglomerate merger will be determined on a rather selective
basis. This view is shared by James M. and D, Jeanne
Patterson, who believe that future court decisions will be
based on actual recognition of the motive or motives under-
lying the merger but will be stated in terms of traditional
arguments.^
^"Justice Agency Issues Merger Guide," The Wall Street Journal, CLXXI (May 31, 1968), p. 4.
^Ibid. ^Patterson, op. cit., p. 43.
39
In any case, the future of conglomerate mergers is
clouded. The existing laws are apparently founded on the
premise that the maintenance of "competition" will best serve
the interests of society, and it is possible, as many writers
point out, that the use of this older concept may be
inappropriate in an environment characterized by modern
technology, human capital, and countervailing power. Perhaps
the current status of the law is best summed up in the
following quotation:
The language of the Clayton Act leaves most of the burden of deciding the issue of conglomerate size up to the courts. It is of course unwise, and probably impossible, not to leave the Supreme Court considerable discretion in such matters if the legislation is to be durable and effective. Still, by its nature the Court is bound to follow the course of its own evolving logic. When it encounters a new situation, it searches among past approaches for appropriate solutions. In the case of the conglomerate merger, the Court has chosen to deal with the problem by extending the traditional struc-tural tests (number of firms and market share) that it developed for vertical and horizontal mergers via the rubric of potential competition. It is not at all clear that such a test can be anything but arbitrarily applied to pure conglomerates. Even more important, it is not clear that the structural test should be applied to the conglom-erate case.9
Possible Efficiencies Released in Mergers
Logic demands that the attention of both investors and
public-policy makers be directed toward possible efficiencies
released in mergers. "Synergy" in financial jargon has come
9Ibid., pp. 47-48.
40
to mean efficiency released through merger. The release of
synergy is the most desirable outcome of a merger since all
interests can be served if this occurs. That is, with the
release of synergy it is possible for management and stock-
holders to benefit through increased profits while society's
resources are being employed more efficiently.
In theory, for synergy potential to exist, the companies
being combined must have similar or complementary needs in
areas such as management, financing, production, technology,
and distribution. Management synergy can be thought of as
the increment of efficiency which results from bringing
together the separate firms under a single management. Like-
wise, synergy release in the areas of finance, production,
technology, and distribution can be viewed as the change in
overall efficiency which results in these functional areas
when the units forming the conglomerate are brought
together. It is, of course, not possible to measure synergy.
There is general agreement that the potential for
synergy exists in many mergers; however, releasing it is
another matter. In the words of one executive, "Don't think
it's hard to release synergy; it's not. It's damn near
impossible."-'-® Almost every survey encountered emphasizes
the importance of management in releasing synergy. The
following quotation is indicative of such emphasis.
•^John Kitching, "Why Do Mergers Miscarry?" Harvard Business Review, XLV (November-December, 1967), p. 94.
41
The element critical for success is not the potential amount of synergy to be released in com-bining two companies. Rather, it is the existence or absence of managers of change—men who can catalyze the combination process. In the most successful mergers, either the acquiring company brought in new managers of change or it motivated the old management to introduce profitable changes.H
Perhaps the most extensive study done to date regarding
synergy was the one conducted by John Kitching, who investi-
gated the experiences of companies two to seven years
following acquisitions. Twenty-two companies were included
in the study, and some rather interesting results were
obtained. The acquisitions of the twenty-two companies were
analyzed and classified in the following manner:
Horizontal—Same industry as buying company, with
approximately the same customers and suppliers.
Vertical integration—Major supplier or customer of the
buying company and in the same industry.
Concentric marketing—Same customer types as buying
company but different technology.
Concentric technology—Same technology as buying company
but different customer types.
Conglomerate--Customers and technology different from
those of the buying company.-^
•^Ibid. , pp. 85 and 91.
l^Ibid.r p. 85; Note: The above definitions are not standard ones--the definitions to be used in this paper are .presented in a subsequent section.
42
D\ie to the nature of synergy, the only quantitative data
which can be formulated is that which is based upon quali-
tative opinions. Accordingly, John Kitching asked the
executives of the various companies interviewed to rate the
degree of synergy attainable in various functions and for
different types of mergers. The following table contains
part of the results of this survey:
TABLE III
SYNERGY POTENTIAL
Type of Merger Finance Marketing Technology Productioi
Conglomerate 100 58 20 32
Concentric Technology
100 72 72 27
Concentric Marketing
100 100 57 72
Horizontal 96 100 41 29
All Categories 100 74 33 36
Source: John Kitching, "Why Do Mergers Miscarry? Harvard Business Review, XLV (November-December, 1967), p. 93,
Executives scored the dollar payoff of synergy as "high,"
"medium," "low," or "none" for each acquisition case. The
scores were converted into arbitrary units, with 100 repre-
senting the score for the highest-rated function. Thus,
the table shows relative payoff values for each function.
43
According to the total ordinal values assigned to each
type of merger in Table III, the synergy potential is least
in the case of conglomerate-type mergers. Also, the ordinal
rankings of synergy potential for the functional areas are,
in this order, finance, marketing, technology, and pro-
duction—a ranking which seems to contradict common sense.
According to John Kitching, synergy potential would osten-
sibly be highest in the case of production, since efficiencies
of scale, quantity discounts, and more efficient machinery
would appear to offer very substantial opportunities for
e f f i c i e n c i e s . - ^ Also, technology would appear to be an area
of high synergy potential because of possible sharing of
expensive R & D results and technological processes. The
sharing of marketing facilities, organizational efficiencies,
and financing abilities would logically rank toward the
bottom of the scale in terms of synergy potential. As the
chart shows, the reverse of this order was suggested by the
executives interviewed.
A possible explanation for the high ranking of financial
synergy and marketing synergy is that, in these areas,
synergy is easiest to release. For example, greater
borrowing potential of the parent company or excess cash "
flow could enable the newly acquired firm to immediately
undertake profitable projects which they were unable to
"^Ibid. , p. 92.
44
finance by themselves. Also, it is suggested that the
relative ease of training a sales force to sell a new product
would make marketing synergies relatively easy to release.
The same executives whose responses are tabulated in
Table III were asked to rank the various categories of
synergy potential in terms of ease of release. Table IV on
page 45 is a summary of the results of their opinions.
According to results tabulated in Table IV, synergy
release is the lowest for conglomerates in all categories
except finance. That is, relative to other types of mergers,
the release of marketing efficiencies, technology
efficiencies, and production efficiencies is harder in the
case of conglomerates.
To conclude, synergy potential and the release of
synergy appears to be highest in the case of mergers which
combine financing and marketing functions. It is possible,
however, that these are the areas in which synergy release
is attained most rapidly and therefore most obviously. In
the long-run, it may be possible that production and tech-
nology synergies play a significant but less immediate role.
Overriding both the long-run and short-run considerations,
however, is the apparent need for "managers of change" (or
management synergy) to release synergy potential in the
various functional areas.
While the results of Kitching's survey are highly
qualitative and impossible to prove, if they are valid, then
45
TABLE IV
SYNERGY RELEASE
Type of Merger Finance Marketing Technology Production
Conglomerate 100 42 18 24
Concentric Technology
65 75 100 65
Concentric Marketing
100 74 21 47
Horizontal 100 67 33 24
All Categories 100 59
_ 11 t.-til .. r\
33 34
L ^ it
Harvard Business Review, XLV (November-December, 1967), p. 93,
Scores are tabulated as noted for Table III.
the current antitrust trend toward favoring "pure"14 con-
glomerates may be the least desirable alternative from an
efficiency point of view. In such mergers the only
functional area of synergy potential is in the financial
area. Such a conclusion, however, assumes that there is no
validity to the public-policy maker's desires to "maintain
competition."
Definition of a Conglomerate Merger
A factor which complicates valid inferences about
conglomerates is the lack of unanimity of definition. A
A definition of "pure" conglomerates is in the fol-lowing section of the text.
46
conglomerate merger deemed a product extension merger may •
involve companies with similar customers, similar technol-
ogies, and similar production processes. Thus, a conglomerate
merger may be pure or mixed, depending upon the degree of
synergy potential in various functional areas.
For discussion purposes, in this paper conglomerate
mergers are classified according to synergy potential as
1• Pure conglomerate merger—one in which no sub-
stantial potential exists for sharing production, technology,
and marketing functions.
Mixed conglomerate merger—one in which some
combination of potential exists for substantial sharing of
production, technology, and marketing functions but not
enough so that the merger could be considered a vertical,
horizontal, or circular one.
Possible Motives of Stockholders and Management Leading to the Growth
of Conglomerate Companies
The objectives and interests of stockholders and manage-
ment do not necessarily coincide. Furthermore, it is never
possible to determine unequivocably the motive or motives
underlying the behavior of individuals or groups of indi-
viduals. Accordingly, the following discussion of motives
and interests is intended only to suggest some of the more
obvious possibilities.
47
The motives of the stockholders of both the acquired
company and acquiring company must generally be satisfied
before a merger can be consummated. The stockholders of the
company being acquired may receive cash, securities, or some
form of property settlement in exchange for their securities.
It is also possible that they could receive some package of
the forenamed types of payment. In all cases, for the
transaction to be rational, a majority of both groups of
stockholders must believe that they are receiving more in
terms of current values than they are giving up.
As the supply of attractive merger candidates continues
to decline in the face of increased demand, the bargaining
position of the acquired companies will also improve.
Accordingly, at some point in time, the motives of the
stockholders of the acquired companies may become at least a
moderate constraint on merger activity.
The following quotation illustrates one of the possible
ways for a parent company's management to satisfy the
desires of the stockholders of both the acquired and
acquiring company:
The purchase of another company has been based as a rule, on an exchange of securities designed to yield tidy benefits to all parties concerned, without much cost in cash to the acquiring company.
Usually, the stockholders of the company acquired are given preferred stock, which they are expected to exchange at some future date for the common stock of the conglomerate if the market value of the stock increases. The preferred,
48
meanwhile, is to pay them a dividend substantially larger than the dividend that they have been getting on the stock of the acquired company.
This preferred dividend, generous as it is, still takes only a portion of the profits the acquired company has been earning. That leaves the balance of its profits to be added to the profits on the common stock of the conglomerate, thus helping it to report a rise in earnings.15
In instances where the stockholders rely heavily on
market performance rather than on fundamental values based
on income potential, the outcome is not always as favorable
as the preceding example might imply. Consider the trans-
action described below.
Borrow several million dollars, using stock in a subsidiary firm as collateral. Buy another company, merge it into the subsidiary and sell additional shares of that subsidiary to the public. You can then pay off the loan while retaining majority control of both the subsidiary and the new firm. Do you recognize the strategy? Of course, it's the merger game—and that is only one of the possible m o v e s . 1 6
In a transaction such as the one just described, it is
possible that the motives of the stockholders of the parent
company will be promoted at the expense of the naive stock-
holder who has bought the new shares in the subsidiary
company. If, however, the market continues to support the
price of the subsidiary company's stock, even the
15"'Glamour stocks' That Ran into Trouble," U.S. News & World Report, LXIV (March 18, 1968), p. 113.
l^James M. and D. Jeanne Patterson, "Conglomerates: The Legal Issues," Business Horizons, XI (February, 1968), p. 42.
49
performance-oriented stockholder may benefit from the trans-
action.
Except in those cases in which management is motivated
by public welfare or power and prestige, their motives can be
expected to coincide rather closely with those of the share-
holders of the company which they manage. That is, any
action by management which is directed toward increased
profits, increased market price of the stock, or increased
stability of earnings could be expected to further the
interests of both groups.
In earlier sections of this thesis it has been suggested
that the attainment of acquisition income may be one of the
dominant motives underlying the growth of conglomerates.
The possibility that profits may be increased because of the
release of synergy has also been suggested as a possible
motive. There is still a third class of motives, which
might be called institutional motives. That is, motives
which are intended to promote overall profits or stability
but are not dependent upon acquisition income or the release
of synergy could be referred to as institutional motives.
An example of such an acquisition is one in which a company
is acquired for its loss carry forward.
In many mergers, more than one class of motive may be
present. The following is a listing of motives which are
primarily institutional in nature but with synergy implied
in several instances:
1. Obtain new products. Industrial annals are full of examples of small companies that became big through acquisitions of a new product or material. When synthetic fibers were perfected, for example, there followed a wave of mergers of companies which had never been in the textile or related business with companies which had mastered the application of these new fibers.
2. Offset declining markets. Innumerable producers of steam locomotive parts and components were in the market for new company acquisitions when the diesel locomotive outmoded steam. A more recent example is the urgent efforts of cigarette companies to get into other fields just in case the anti-smoking campaign is successful.
3. Stimulate growth. In this era of economic expansion, it's a rare company that is satisfied to stand still. It sees in mergers a technique for keeping in step with the nation's growth. Thus a maker of hand tools (hammers, saws, screw drivers, etc.) may not be content with a market that is only mildly expansive. It may well try to acquire a power tool subsidiary as a means of cashing in on the homemakers' love affair with make-it-yourself.
4. Stabilize seasonal and cyclical trends. In the old days, the coal dealer bought up the local ice dealer to insure summer and winter business, Today, a more likely example is the refrigerator maker who takes on a line of space heaters and humidifiers to assure all-year-round demand for what he makes.
5. Put excess working capital to work. This is a common and understandable reason for many mergers. Idle capital can be better utilized in investment in a new firm than gathering dust in a bank.
6. Convert a holding company to an operating company. Tax considerations often stimulate a private holding company to convert into an operating firm. The most painless way to do it is to buy up an existing and going operation.17
l^johan Bjorksten, "Merger Lemons," Acquisitions and Mergers, I (Fall, 1965), p. 37.
51
The following diagram is perhaps appropriate for
summarizing the range of possible motives underlying the
growth of conglomerate mergers:
Financial Manipulation—main moti-vation is attainment of acquisition income
2. Institutional Motives—main moti-Merger vation is attainment of increased Spectrum stability or profits by means
other than acquisition or through the release of synergy
3. Synergy Motive—main motivation is increased profits through the release of synergy
Fig. 6—-Merger Spectrum
Moving down the merger spectrum from one to three, the
degree of desirability probably increases from the viewpoint
of both investors and public-policy makers. In particular,
it would be desirable to isolate those companies which rely
solely on financial manipulation at the expense of efficiency,
Unfortunately, existing theories of value and methods of
analysis are inadequate for making such a distinction.
CHAPTER IV
THEORIES OF VALUE AND THEIR IMPLEMENTATION
In the last chapter it was pointed out that a range of
possible objectives and efficiencies may underlie the growth
of companies engaged in conglomerate mergers. Unfortunately,
investors do not have appropriate methods for placing values
on the shares of these companies relative to their income
potential for the following reasons:
1. The lumping together of operating and acquisition
income in financial reports,
2. The absence of established earnings trends under
various economic conditions,
3. The data regarding financial performance are gene-
rally provided only on a consolidated basis,
4. The lack of uniformity in accounting methods between
companies and before and after merger,
5. The lack of comparability of operating circumstances
before and after merger,
6. The presence of diversification of earnings due to
multi-industry orientation.
Since the factors listed above preclude the application
of traditional methods of valuation, a new method and new
approach are imperative. In pursuit of this objective, the
52
53
first part of this chapter is concerned with theories of
value in general and the various alternative theories which
are possible. An effort is made to show that theories of
value are not theories in a scientific sense, but rather
that they are decision models which have objectives and
assumptions implicit in them. Finally, in the latter part
of this chapter, a theory for valuing the stock of conglom-
erate companies is developed which gives recognition to
problem areas two and six listed above.
The chapters which follow this one are concerned with
the .remaining problem areas and the testing of some hypotheses
which relate directly to the application of the theory to a
selected group of conglomerates. Also, the final development
and refinement of the theory of value in question is con-
summated.
Theories of Value
When a quantification is sought for a characteristic
such as investment value or speculative value, the quantifi-
cation is attained through valuation processes rather than
measurement processes. Characteristics such as investment
value and speculative value cannot be measured because they
are subjective characteristics which exist only in the minds
of those placing the values, and, accordingly, an element of
choice exists regarding the method in which numerical values
are assigned.
54
Stockholders have historically exerted their choice
prerogative in the theories of value and valuation criteria
which they have applied. In his recent article in the
Financial Analysts Journal, "Performance~-the Latest Name
for Speculation?", David Babson described the evolution in
appraisal emphasis as historically moving through the
following states:
1. Net worth, book value and physical assets to
2. Income return, dividends and yield to
3. Earnings and earnings reliability to
4. Long-range growth rate of earnings and now to
5. Instant earnings growth.1
The last appraisal method mentioned above is an obvious
reference to stockholders who are motivated to buy shares
in conglomerate companies because of the ability of such
companies to show instant increases in income through
acquisition. Perhaps it should be pointed out, in fairness
to investors, that their reliance on instant earnings growth
may be a by-product of their inability to discern long-run
income potential.^ Also, it is perhaps important to note
that the range of possible valuation criteria is not limited
to five or six, but is virtually unlimited.
-'-Abraham J. Briloff, "Distortions Arising from Pooling-of-interests Accounting," Financial Analysts Journal, XXIV (March-April, 1968), p. 75.
^John M. Keynes, The General Theory (New York, 1935), p. 148. ~
55
Investment and Speculative Theories of Value
While the actual number of motives and theories of value
which could be applied in security purchases are nearly
unlimited, there are only three formalized approaches.
The three approaches referred to are those of the fundamen-
talists, the technicians, and the random-walk theorists.^
Fundamentalists attempt to define a type of intrinsic
value for particular securities based upon long-run economic
expectations, whereas random-walk theorists and technicians
attempt to make projections of market prices. By either
making assumptions about what creates future income or what
causes market prices to change, the three approaches are
used to develop guides to investment or speculative decision
making.
There is a tendency to treat the three schools of
thought as being totally unrelated. This tendency has caused
a certain amount of rigidity and intolerance among advocates
of the different point of view. In reality, all three
approaches are characterized by decision models whose real
differences can be attributed to differing objectives and
differing assumptions.
For discussion of fundamentalists' approach to valu-ation see Benjamin Graham and others, Security Analysis (New York, 1962); for differentiation between fundamental and technical approach to security valuation see John C. Clendenin and George A. Christy, Introduction to Investments, 5th edition (New York, 1969), p. 254^
56
The following section•in this chapter demonstrates that
theories of value are equivalent to decision models, and also
a detailed examination is made of the elements of decision-
making situations. This is done in an effort to create the
proper background for developing a theory of value which is
applicable to conglomerate companies.
Decision Models
It is possible to state, in a somewhat unrewarding
manner, that an investor having a certain amount of funds
available for investment will rationally divide his funds
among alternative investments in such a way that
Expected benefits Expected benefits Expected benefits from the purchase from the purchase from the purchase of one more share of one more share of one more share of Company X = of Company Y = of the last alter-Price of shares Price of shares native company .in Company X in Company Y Price of shares
in last company
In other words, it is possible to state that a rational
investor will seek to spread his investment funds among
alternative outlets in such a way that the total utility per
dollar invested is a maximum.
In order for a particular investor to approximate the
equality expressed above, methods must exist which are accep-
table for placing values on the alternative investments.
Stated differently, investors are decision makers who are in
a position of having to select the most desirable alternatives
available in light of their own objectives and needs. To
57
select the most desirable alternative, the investor in
question should have a clear understanding of his own objec-
tives and the circumstances surrounding the investment (i_.e.,
a clear understanding of the decision-making situation).
Decision-Making Situations
In order for a decision-making situation to arise there
must exist (1) a decision maker with an objective or problem,
(2) a certain number of factors related to the objective or
problem which are not within the control of the decision
maker, and (3) a certain number of alternatives of action.
Without all three of these elements, a decision-making
situation cannot exist. For discussion purposes those
factors which are not within the control of the decision
maker are referred to as states of the world.
While it is not always obvious, a relationship exists
between the problem or objective of the decision maker and
the particular state of the world surrounding the decision.
The truth of this proposition can be readily visualized if
one observes that the state of the world surrounding the
decision is responsible for the problem or objective faced
by the decision maker. In other words, if there were not
resistance in the form of environmental circumstances, then
problems or objectives would not exist. The decision maker
faced with a set of circumstances attempts to take an action
which will lead to a different state of the world.
58
Hopefully, this subsequent state of the world will correspond
to the one corresponding with preconceived objectives.
The elements of a decision-making situation are perhaps
best visualized in matrix form. Possible states of the
world are typically listed in columns, and alternatives
listed in rows. In the following matrix the Y's denote
possible states of the world, the X's refer to possible
alternatives of action (alternative investment outlets), and
the A's refer to the expected outcomes of the various alter-
natives, given the corresponding states of the world.
Y1 *2 y3 *4 Y5
X1 A11 a12 >
t-4 u>
1
a14 a15
*2 a21 A22 a23 a24 A25
*3 A31 A32 A33 a34 A35
X4 A41 a42 a43 a44 A45
X5 A51 a52 a53 a54 A55
Fig. 7—Decision matrix under conditions of future uncertainty.
In the above matrix, five possible states of the world
are represented, along with five possible alternatives of
action (alternative investments). In theory, the number of
rows and columns could be expected to include as many
59
alternatives and states of the world as are considered pos-
sible. The A's are a measure of the expected outcomes of
the various alternatives, given a specific objective and
values for states of the world. Accordingly, the A's can be
considered an ordinal measure of the value of an alternative,
given a specific state of the world and an objective.
In order to fill in values for the A's in the payoff
matrix, a relationship must be derived between alternatives,
states of the world, and an objective variable "0," (The
nature of this relationship will be explained more fully in
the following paragraph.) In mathematical terms, the A
variable is the dependent variable, and the 0 and Y variables
are the independent ones. Accordingly, the relationship
between values for alternatives, objectives, and states of
the world can be expressed as
A = f(0,Y)
Under conditions of certainty (assumed certainty) about
states of the world surrounding the decision, it is not
necessary to describe the decision-making situation in matrix
form. The desired values for the objective variable and
values for states of the world can be substituted directly
into the decision model, and a corresponding value for A is
determined for each alternative under consideration. Under
conditions of certainty the resultant values for A are
adequate for decision-making purposes. That is, the ordinal
60
rankings which result for the A's allow the most desirable
alternative to be selected.
The theories of value followed by fundamentalists are
decision models of the above form. That is, the appropriate
numerical valuation for a security is determined by (1)
specifying an objective in numerical terms, (2) deriving
numerical values for the state of the world under conditions
of assumed certainty, and (3) solving the decision model
(i_.e., theory of value) for the corresponding value of A,
the alternative of action. Thus, under these conditions,
the appropriate valuation for the security is also the appro-
priate numerical value for the alternative of action under
the conditions prescribed.
The basic valuation model of the fundamentalists is the
one which prescribes the present value of future dividends,
given a desired rate of return and a projection of future
dividends. The model is usually expressed in this form:
t " k P = \ D (1 + G)fc
/," (1 + R)t~
t = 1
In the above model D is the current dividend rate, G is the
percentage growth rate of dividends, t is time, R is the rate
at which the future dividends are discounted, P is the
present value of the future dividends, and k is the length
of time for which dividends are to be received. Note that
D, G, and t are environmental circumstances, P is the value
61
for the alternative of action, and R is the objective
variable. By prescribing values for the states of the world
and also for the objective variable, the corresponding value
of P is also determined.
While some of the models used by fundamentalists are
more elaborate than the one presented above, all of them
follow the same form. That is, they all have environmental
circumstances and an objective variable as independent
variables. Also, the dependent variable is always the
numerical valuation for the security being valued, which is
also a numerical valuation for an alternative of action.
Thus, to summarize, it can be shown that so-called
theories of intrinsic value are nothing more than decision
models whose dependent variable is the derived numerical
valuation for the security. Also, it can be shown that such
decision models generally assume conditions of certainty
about the state of the world. A legitimate question to be
asked at this point is—-Why have theories of intrinsic value
been limited to such simplified assumptions about the nature
of environmental circumstances? The next section in this
chapter discusses and presents the different possible
decision-making situations in terms of the manner in which
the states of the world are assumed to be generated.
Accordingly, if such a classification scheme is all-inclusive,
then all possible methods of valuing securities will also
be enumerated.
62
Types of Decision-Making Situations
According to the foregoing analysis, in order for a
decision-making situation to have any meaning, the decision
maker must have some preconceived notion about the proper
response he should make whenever the state of the world
assumes a particular configuration. Assuming that the
reaction of the decision maker in response to his environment
is carefully conceived (i.-e. f derived under the assumptions
of appropriate objectives, and by making valid assumptions
about the relationship between alternatives, states of the
world, and objectives), then the major problem faced by the
decision maker is determining values for the variables that
describe the state of the world.
Decision-making situations can be classified according
to what is known about the state of the world. In many
instances the assumption is made that the state of the world
is known with certainty. This is always an assumption,
however, because perfect knowledge is never possible. In
many instances the state of the world can be identified with
sufficient accuracy to warrant its being considered as
certain. Probably the most common reason for lack of
reasonable certainty is that decisions must be made before
the relevant state of the world materializes. For example,
in financing, production, and marketing decisions, one of
the major variables describing the state of the world is
the demand for the goods which are being financed and
63
produced. Since the demand will not be known until a future
date, estimates must be made beforehand.
Another type of uncertainty can exist in a decision-
making situation because of physical or financial limitations
of the decision maker to perceive existing facts about the
state of the world. If, for example, a decision maker
needed to know the average age of the citizens in a particular
community at a particular point in time, both physical and
financial constraints could prohibit such a measurement. It
is significant that the age of people at a particular point
in time is a fact. Also, such facts are indicative of a
state of the world which has already materialized, and the
uncertainty which exists is fundamentally different from the
type previously described.
For discussion purposes, the first type of uncertainty
described above is referred to as "future" uncertainty, and
the latter type is referred to as "knowledge" uncertainty.
To reiterate briefly, future uncertainty exists because the
state of the world has not materialized at the time the
decision must be made, and furthermore the forces acting upon
the environment are not adequately understood so that
accurate predictions can be made about the future. The
second type, or knowledge uncertainty, exists because of the
decision maker's inability to attain a valuation for state
of the world variables which exist at the time the decision
is to be made. Thus, decision-making situations can be
64
broadly classified as those under certainty and those under
uncertainty. Within the "uncertainty" category two addi-
tional subcategories are discernible. The following diagram *
is intended to be indicative of these broad divisions.
TYPES OF DECISION-MAKING SITUATIONS \
CERTAINTY UNCERTAINTY ^ X
KNOWLEDGE FUTURE UNCERTAINTY UNCERTAINTY
Fig. 8--General types of decision-making situations
Within the category of future uncertainty there are
several additional subcategories of decision-making situ-
ations. These two subcategories are defined on the basis of
the nature of the process generating the states of the world,
as either consistent or inconsistent processes. If a con-
sistent process is generating the states of the world then,
at least in terms of probability, there is order to the
decision environment. If, however, there is an inconsistent
process generating the states of the world, then no probabi-
listic order other than subjective probabilities can be
assigned to the generation of the states of the world. The
completed decision diagram is shown on page 65.
Figure 9 is helpful in identifying different decision-
making situations. As previously noted, the diagram is
based upon the assumption that decision-making situations
can be meaningfully classified according to the nature of
65
TYPES OF DECISION-MAKING SITUATIONS
CERTAINTY UNCERTAINTY
KNOWLEDGE FUTURE UNCERTAINTY UNCERTAINTY
CONSISTENT INCONSISTENT PROCESS PROCESS
Fig. 9—Specific types of decision-making situations
the process generating the independent variables
environmental circumstances) of the decision model. Also,
the same format is quite useful for classifying various
types of mathematical techniques which exist for solving the
corresponding decision-making situation for the desirable
alternative. The diagram presented in Figure 10 represents
such a classification scheme.
To summarize, it can be demonstrated that the theories
of investment value used by fundamentalists are really
nothing more than decision models of the form described in
this chapter. Since this is true, it is logical to suggest
that such theories have been limited to rather simplified
assumptions about how the relevant states of the world are
being generated. Z lso, it appears likely that fundamen-
talists have not availed themselves of some of the more
sophisticated techniques used for solving decision-making
situations.
66
DECISION-MAKING SITUATIONS
CERTAINTY
KNOWLEDGE UNCERTAINTY
UNCERTAINTY
FUTURE UNCERTAINTY
CONSISTENT PROCESSES
METHODS OF ATTACK
1. Statistical inference
1. Statistical inference
2. Game theory 3. Simulation
1. Bayesian decision theory
1. Max. & Min. (Calculus)
2. Linear programming
3. Simple substitution
4. Math programming
Fig. 10--Types of decision-making situations and methods of attack.
INCONSISTENT PROCESSES
The selection of the most appropriate method for making
investment decisions / the selection of a theory of
value) depends largely on whether or not proper recognition
is given to the method in which the relevant states of the
world are being generated. The following section of this
chapter discusses the assumptions underlying current
approaches to investment decisions and suggests why such
assumptions may not be appropriate for valuing investment
alternatives involving conglomerate companies.
67
Generation of the Relevant States of the World
According to the format developed in the previous
section, four different patterns exist in which numerical
values for states of the world can be generated. These four
patterns are (1) complete certainty, (2) knowledge uncer-
tainty, (3) future uncertainty with a consistent process
generating the values, and (4) future uncertainty with an
inconsistent process generating values for the independent
variables.
In the first of the categories listed above, it is
assumed that the values exist at the time the decision is to
be made, and, furthermore, such valuations are assumed to be
known with certainty. In the second instance, the relevant
state of the world exists but for one or more reasons is
difficult or impossible to ascertain. In the third type of
situation, the relevant state of the world has not been
generated but can be anticipated with some degree of reli-
ability because of order in the underlying process. An
example of this type of process would be a production process
which produces acceptable and unacceptable products in
accordance with a Bernoulli process.^ Finally, in the latter
category, are events which have not yet occurred and to which
no dependable order exists in their materialization.
^For explanation and analysis of Bernoulli processes, see pp. 29--32 of Quantitative Analysis for Business Decisions, by Harold Bierman and others (Homewood, Illinois, .1965) .
68
What then is the nature of the force or forces which
generate values for theories of value used by fundamentalists?
If such forces are inconsistent, then perhaps the assumption
of certainty is not justified.
Basic Models Used by Fundamentalists
Underlying most valuation models currently in use is
the concept of rate of return on investment. Depending upon
the particular purpose, writers may emphasize either earnings
per share or dividends per share; also, allowances for growth,
.inflation, and risk can be incorporated into the individual
models. Theories based on the concept of current yield
generally assume a form similar to the following:
Earnings per Rate of Return = Period
per Period Investment
i = = P
The relationship expressed in the preceding equation is a
truism when applied to variables of past experience. When
applied in this manner the values for the variables are known
with reasonable certainty. However, as a theory of value a
standard £ ratio is assumed—depending upon the rate of E
return desired (i = £). When used as a theory of value, the x E
investment value is determined by the product of the standard
multiple times expected earnings. Used in this manner, E is
69
a future environmental circumstance and is derived under
conditions of future uncertainty. Furthermore, there is
little evidence to suggest that orderly or consistent forces
are responsible for producing consequent values for E.
A slightly more complicated model can be derived by
extending the concept of rate of return on investment to
include more than one time period. Starting with the same
basic equation,
i _ E _ S - P
and letting S denote total investment value at the end of a
time period while P refers to the beginning principal, the
equation can be rewritten this way:
S = P (1 + i)
When extended to more than one time period (^.e., to n time
periods), the equation assumes the following form:
S = P (1 + i)n
Once again, when the equation is applied to known values for
i, n, and P, no special difficulties arise; however, when
used as the basis of a theory of value, P becomes the
dependent variable, and S becomes one of the independent
variables. Under these conditions S is valued under con-
ditions of future uncertainty, and no evidence suggests that
consistent processes are generating such values. When
70
solved for P, the above equation is the one underlying all
"present value" theories of value.
It is possible to combine both current yield and yield
to maturity concepts into one theory of value. This is done
by almost all writers who desire to develop methods for
valuing common stocks. One such writer is Malkiel, who set
forth the following model in The American Economic Review,
for December, 1963.5
p = D (1+g) + p (1+g)2 + p (l+g)n + MsE(l+g)" (1+r) (l+r)2 (l+r)n (1+r)1"*
Where
D = Current dividends in dollars
Ms = Standard earnings multiplier S & P averages
g = Growth rate of company as a percent per year
E = Current earnings in dollars
r = Standard rate of growth S & P averages or the apparent marginal efficiency of the representative standard share
n = Number of years forecasted
P = Present value of future stream of receipts.
Malkiel's model makes provision for growth and for
value of the stock in the n'th year. The last term of the
model is based upon a current yield concept in the form of a
standard P multiple which is discounted, along with all E .
Paul F. Wendt, "Current Growth Stock Valuation Methods," Financial Analysts Journal, XXI (March-April, 1965), p. 91, citing B. G. Malkiel, The American Economic Review (December, 1963). ~ "
71
preceding dividends, to a present value. The implementation
of this model, like the simpler ones mentioned previously, I
requires that projections of future states of the world be
made.
While a number of other forms of models exist besides
the ones previously presented, they all are based upon the
same present value notion, and all require the projection of
future earnings or future dividends in their implementation.
Furthermore, as noted previously, there is no evidence to
support the belief that future earnings or future dividends
are being generated, in all cases, by a consistent process.
J. M. Keynes was probably among the first to point out
the difficulties inherent in any attempt to make long-range
projections of fundamental economic behavior. Our knowledge of the factors which will
govern the yield of an investment some years hence is usually very slight and often negligible. If we speak frankly, we have to admit that our basis of knowledge for estimating the yield ten years hence of a railway, a cooper mine, a textile factory, the good will of a patent medicine, an Atlantic liner, a building in the City of London amounts to little and sometimes to nothing; or even five years hence.6
In a subsequent passage, Keynes comments about the effects
of future uncertainty on the valuation processes or theories
of value.
It might have been supposed that competition between expert professionals, possessing judgment and knowledge beyond that of the average private
^John M. Keynes, The General Theory (New York, 1964), p. 149.
72
investor, would correct the vagaries of the ignorant individual left to himself . . . most of these persons are, in fact, largely concerned, not with making superior long-term forecasts of the probable yield of an investment over its whole life, but with foreseeing changes in the conventional basis of valuation a short time ahead of the general public.7
This battle of wits to anticipate the basis of conventional valuation a few months hence, rather than the prospective yield of an investment over a long term of years, does not even require gulls amongst the public to feed the maws of the pro-fessional;—it can be played by professionals amongst themselves. Nor is it necessary that anyone should keep his simple faith in the conven-tional basis of valuation having any genuine long-term validity. For it is, so to speak, a game of Snap, of Old Maid, of Musical Chairs—a pastime in which he is victor who says Snap neither too soon nor too late, who passes the Old Maid to his neigh-bour before the game is over, who secures a chair for himself when the music stops. These games can be played with zest and enjoyment, though all the players know that it is the Old Maid which is circulating, or that when the music stops some of the players will find themselves unseated.8
In this latter paragraph, Keynes suggests that the
Technical approach to security valuation is a natural corol-
lary to the existence of a market characterized by future
uncertainty.
In a decision-making context, it is possible to state
that Fundamentalists and Technicians have theories of value
which are based upon slightly different objectives and
different assumptions about the related states of the world.
7Ibid., p. 150.
^Ibid.
73
Fundamentalists attempt to find a security selling below its
investment value, whereas Technicians, or Chartists, attempt
to find securities whose price will rise in the near future.
The theories of value used by Technicians appear most fre-
quently in the form of charts or graphs—the basic assumption
seems to be that patterns exist in the way that average
opinion places values on securities, and by being familiar
with such patterns, one can anticipate price changes.
Conflicting Theories of Value
The objective underlying Technical theories of value is
to anticipate price changes, and the alternative dictated by
such theories is to buy or sell the security, depending upon
the direction of expected price changes. Generically
speaking, this type of objective and method of selecting an
alternative is quite different from the corresponding ones
used by Fundamentalists.
In addition to differing objectives, one of the main
differences between the theories of value used by Technicians
and those of Fundamentalists is that Technicians' models have
existing states of the world as independent variables rather
than future states, as is true in the case of Fundamentalists.
Thus, the models used by Technicians avoid the problem of
future uncertainty by assuming that a consistent process
generates price movements, and by assuming that if one has
knowledge of the pattern, then all one has to do is plug
74
in the relevant state of the world today in order to predict
tomorrow's prices.
In all fairness to both the Fundamentalists and Tech-
nicians, spokesmen for both schools generally admit that
their conclusions and forecasts are subject to some doubt,
and in the case of the majority of Fundamentalists, a pro-
vision is usually made for an appropriate margin of safety
or for diversification. However, the basic approaches of
both schools remain tantamount to ignoring the possibility
that an investment in a single security is a commitment to
the future which, in turn, may b& subject to largely im-
perfect or unknown forces.
In light of the conflicts and controversies between the
views of the Fundamentalists and those of the Technicians,
it is not surprising that studies have been conducted whose
conclusions support the random-walk thesis. The best-known
of such studies is probably the one done by Eugene Famma.^
The conclusions of this study seem to imply that both the
Fundamentalists and the Technicians are unable to determine
information or theories of value which will allow decisions
to be made that yield profits greater than could be attained
by a naive buy-and-hold strategy.
Thus, at one extreme both the Fundamentalists and the
Technicians argue that there is order to the generation of
^Eugene Famma, "Random Walks in Stock Market Prices," Financial Analysts Journal, XXI (Sept.-Oct., 1965), pp. 55-59,
75
future events, but they disagree as to the nature of the
order. At the other extreme, proponents of the random-walk
thesis argue that stock price movements are sufficiently
random so that the average investor can do no better than a
buy-and-hold strategy.
Thus, to summarize these three different points of view,
the following diagram is used once more:
DECISION-MAKING SITUATIONS
CERTAINTY
Fundamentalists
KNOWLEDGE UNCERTAINTY
UNCERTAINTY
FUTURE UNCERTAINTY
PERFECT PROCESS
Technicians and
Fundamentalists
IMPERFECT PROCESS
Random Walk
Fig. 11—Assumptions underlying the major approaches to security valuation.
Of course, the Fundamentalists have basic economic
variables as independent variables, whereas Technicians have
present price and volume movements. Thus, the order assumed
by Technicians is not orderly generation of economic con-
ditions, but rather, they assume order exists in the methods
which individuals follow, on the average, when placing values
76
on securities. Finally, Random-walk theorists refuse to
recognize dependable order in either instance.
The Decision Environment Surrounding Conglomerate Companies
One of the critical elements in selecting a theory of
value is the assumption about how the states of the world
are being generated. As noted in the previous section, most
of the formalized theories which exist have avoided the
problem of future uncertainty by either assuming certainty
or by assuming that a perfectly consistent force is gene-
rating future events. Or, as in the case of Random-walk
advocates, no dependable order is recognized at all.
Regarding the valuation of conglomerate companies, it
may be a bit premature to ask if future uncertainty exists
when the present valuation environment is clouded by the
following factors:
1. The lumping together of operating and acquisition
income in financial reports.
2. The absence of established earnings trends under
various economic conditions.
3. The data regarding financial performance is
generally provided only on a consolidated basis.
4. The lack of uniformity of accounting methods between
companies before and after merger.
5. The lack of comparability of operating circumstances
before and after merger.
77
6. The presence of diversification of earnings due to
multi-industry orientation.
Because of all of the reasons listed above, but mainly
because no earnings trends exist for conglomerate companies
which include periods of adverse business conditions, the
recognition of uncertainty seems to be an unavoidable conse-
quence.
If uncertainty is recognized, then because the states of
the world underlying the valuation of conglomerate companies
( from the Fundamentalist's point of view) are future income
levels, the particular type of uncertainty to be dealt with
is future uncertainty. Furthermore, if valuation is to
occur at all, there must exist some degree of order to the
generation of the future states of the world—order in the
sense of a probabilistic ordering). Accordingly, the states
of the world applicable to the valuation of conglomerate
companies would appear to be most appropriately considered
as being subject to future uncertainty with something less
than a perfect process generating values for the states of
the world. Under conditions such as these, the diagram
presented in Figure 10 suggests that by default the only
logical choice of method for placing values on alternatives
is with the techniques of Bayesian decision theory. In line
with this reasoning, one of the more common strategies in
Bayesian decision theory is selected as a basis for
developing an appropriate theory of value. Before this is
78
done, however, problems which arise when uncertainty is given
recognition are discussed in the following section.
Problems Introduced with the Recognition of Uncertainty
When emphasis is shifted from conditions of certainty
about the states of the world to conditions of uncertainty,
probabilities are given specific recognition in the theory
of value. Accordingly, each alternative in a decision-making
situation has an outcome—given a desired value for an
objective variable and for states of the world—and a
probability for that outcome. This means that not just one
number but two numbers, both an outcome and a probability
of that outcome, can be associated with each alternative for
each state of the world that is considered to exist. The
contrast between decision making under conditions of cer-
tainty and uncertainty is highlighted by the following two
matrices.
Certainty Uncertainty P rl Pr2 Pr3 Pr4
Y1 Y1 y2 Y3 Y4
Alternative 1 A^ Alternative 1 A11 A12 a13 A14
Alternative 2 A2 Alternative 2 a21 a22 a23 a24
Alternative 3 A3 Alternative 3 a31 a32 a33 a34
Alternative 4 A4 Alternative 4 a41 a42 a4 3 A 4 4
Alternative 5 A5 Alternative 5 A51 a52 a53 a54
Fig. 12—Matrices reflecting the two major types of decision-making situations.
79
As the preceding matrices show, under conditions of
uncertainty, the valuation for each alternative is dependent
upon both conditional outcomes and probabilities. This
means that before a final valuation can occur, the decision
maker must know what his preferences are regarding combinations
of conditional outcomes and probabilities. Thus, a theory
of value must be developed which recognizes both of these
elements.
Objectives Underlying the Valuation of Conglomerate Companies
In order to define a theory of value which is appro-
priate under conditions of uncertainty, one must first
specify the objective upon which the theory is to rest. That
is, a theory of value places values on alternatives relative
to a given value for an objective variable.
It cannot be proved that one objective is better than
another except relative to other objectives and other
assumptions. Accordingly, the selection of an objective is
a personal matter, and if issue is taken with the objective
implicit in a theory of value, then the theory of value is
not appropriate.
The arbitrary nature of the selection of an objective
is evident in the manner in which John Burr Williams, the
father of investment present-value theory, justified his
selection:
80
This definition for investment value which we have chosen is in harmony with the time honored method of economic theory which always begins its investigations by asking, "What would men do if they were perfectly rational and self seeking?"-'-®
Following the lead of Williams, as he appeals to common
sense as authority for his theory, the objective which is
assumed to underlie the rational purchase of common stock
under conditions of future uncertainty is the maximization of
expected present value per dollar invested. That is, it is
assumed that a rational investor will always try to select
the alternative investment having the highest probability-
weighted average of present values—where probabilities and
present values exist for each state of the world considered
possible.
In a decision-making context, it is possible to state
that a decision maker with the objective defined above will
continue to buy shares in Company X only so long as the
following ratio is higher for shares in that company than
the same ratio calculated for shares in another company:
Expected Present Value of Future Income Price of Shares in Company X
The initial intent of examining and specifying an ob-
jective for valuing common stock of conglomerate companies
'^John Burr Williams, Theory of Investment Value (Cambridge, Massachusetts, 1938), pp. 5-6.
81
under conditions of future uncertainty was to develop a cor-
responding theory of value which would give recognition to
both probabilities and conditional present values. Such a
theory is implied by the objective just defined. However,
before developing the theory further, several assumptions
about utility which are implicit in the objective are
examined.
Assumptions about Utility Implied by the Objective to Maximize Expected Present Value
By assuming that a rational investor attempts to maxi-
mize expected present value under conditions of future
uncertainty, several assumptions have been implied regarding
the rational investor's preferences for various expected out-
comes and their related probabilities. To begin with, the
objective variable is the one which transforms values for
expected outcomes and probabilities into a valuation for the
alternative under consideration. Accordingly, if the objec-
tive variable is defined and applied appropriately, the
resultant values for alternatives should correspond precisely
to the preferences of the individual applying the decision
model. Since the value for "r," the expected rate of return,
is considered to be a constant when applied in determining
the expected present value, it follows that one assumption
underlying such an objective is that the individual investor
applying the model will have a constant objective, regardless
of the patterns and amounts which are attributed to expected
82
outcomes and their corresponding probabilities. This means
that an investor is assumed to be indifferent as to the indi-
vidual outcomes and probabilities which create a particular
expected present value so long as the probability-weighted
average of present values is the same. The graph below is
indicative of this implied relationship.
E pvl
E pv2
Pr n
E pvm
a constant level of expected present value (since r is considered to be constant, this is also a constant level of utility)
a constant level of expected present value (utility), which is higher than EpV^
the probability of state of the world "n," which corresponds to present value Pn
a constant level of expected present value "m," where m can take on values corresponding to different levels of indifference
Epvm ~ Prn pn
or
E pvm f(Prn * Pn)
dE = Je Pvm Jpi"
pvm dPrn + jEpVm ^pn
n JPn
P.. and JEpVra ' P m ri j pn
Fig. 13~-Indifference curves reflecting constant expected present value.
Along a single indifference curve (EpV]_ or EpV2, for
example) dEpVm = 0. Thus
83
0 = (Pn)dPrn + (Prn)dPn
Dividing both sides of the equation by dPn, the marginal rate
of substitution is
a p r n = - p r n dPn ?n
The condition implied by the expression above is that
in order for a constant level of utility to be maintained,
a change in one variable must be off-set by a change in the
opposite direction of the other. This is merely a mathe-
matical formulation of the assumption noted previously that
underlies the objective which has been defined.
A Theory of Value for Valuing the Stock of Conglomerates
Having defined the objective which is assumed to charac-
terize the rational purchase of common stock under conditions
of future uncertainty and having presented the assumptions
about utility implied by the objective, the groundwork has
been laid for developing the details of the theory of value
desired.
Keeping in mind that the objective which has been
assumed is the maximization of expected present value, the
theory of value desired is perhaps best viewed as the method
by which a single alternative in a decision-making situation
is assigned a numerical value relative to this objective.
A single alternative under conditions of uncertainty would
84
appear similar to the following:
Alternative X = Pr- Pr^ Pr^ Prn
^1 £*2 ^ 3
In the matrix above, the present value for state of the
world "n" is denoted by the following:
t = s pn = V ~ E t (1 + 9t>fc
(1 + r)t t = 1
In the preceding equation, Et is current earnings per share,
g . is the expected growth rate of earnings per share, t is
time, s is number of time periods, and r is rate of return
demanded per time period.
Expected present value is the theory of valu^ which is
used to value alternative X; it is the probability-weighted
average of the present values under the various states of
the world considered possible, where Prn is the probability
of the state of the world n.
Epv = P1 Prl + p2 * P r2 pn Prn
where
P rl + P r2 P rn = 1
or
n = m t = s
E Et {1 + W
t
P V — (1 + r)t Pn n = 1 t = 1
Alternative_ _ Valuation ~
85
Objective States of Probabilities of Variable ' the World' States of the World
EpV is the theory of value which is appropriate for
placing values on various alternative investments under con-
ditions of future uncertainty. Note that the main difference
between the certainty model and the uncertainty model is
that probabilities are now recognized as independent vari-
ables.
In the preceding equation, EpV is expected present
value, and g-j-n is the expected growth rate in earnings per
share if state of the world "n" occurs. Also, r is the
expected rate of r e t u r n . I n applying this theory of value,
different states of the world considered possible would be
evaluated to determine their likely effect on g^, and a
probability for each state of the world would be derived.
Once g^ and Pr are determined for each state of the world,
then an expected rate of return is specified by the investor.
Given the current level of earnings per share E^ and the
values for all of the other independent variables, then an
expected present value can be derived.
For purposes of valuing the common stock of conglomerate
companies, the assumption is made that the expected present
value model is appropriate. The selection of this model is
based on several considerations. First, the lack c>f ex-
perience of conglomerates under adverse business conditions
-'-•'•See page 86 for definition of r.
86
precludes the practice normally followed by Fundamentalists
of projecting past earnings trends into the future.
Accordingly, the recognition of uncertainty of future income
is hardly avoidable. Second, present value theory is in
conformity with the conservatism which should guide invest-
ment decisions. That is, in a financial atmosphere such as
today's, the ability of an investor to place values on
securities relative to income potential is extremely impor-
tant unless the game of Snap is the real objective motivating
stock purchases. Finally, the expected present value
criterion is conveniently applied to conglomerate companys'
shares because it makes automatic adjustments for risk and
for the effects of diversification of income sources. That
is, by assuming different states of the world when applying
the model, a decision maker builds into the calculation of
expected present value the effects of risk and diversifi-
cation on income. Accordingly, the value for "r" used in
the theory of value is the rate of return demanded exclusive
of the risk factor. This attribute of the model is due to
the recognition of future uncertainty in the model itself.
While the final refinements of the theory proposed are
contained in the last chapter, it seems fitting to present a
simplified example of how the theory might be applied. For
example, assume that five states of the world are possible
-"•2Keynes, op. cit., p. 155.
87
(each characterized by a different gt), and assume that the
present value of the security being valued is calculated as
$95, $106, $75, $65, and $40 for each of the five states of
the world considered possible. Furthermore, let it be
assumed that the probability of state of the world number one
(corresponding to the present value of $95) is one-half, and
the other four states of the world all have a probability of
one-eighth. Under these conditions, the expected present
value is
E p y = (.5) (95) + (.125) (106) + (.125) (75) + (.125) (65) + (.125) (40)
or
E = $33.22 pv
To summarize, the main intent of this chapter has been
to point out the equality of theories of value and decision
models used to place values on alternatives of action under
various conditions of the decision environment. Once the
equality was pointed out, it was suggested that the tra-
ditional assumptions of certainty and perfect processes are
not justified in the valuation of the common stock of con-
glomerate companies. Finally, the basic framework of a
theory of value was developed which recognized future uncer-
tainty and made allowances for the effects of risk and
diversification of income sources. The theory developed
employed one of the simple decision strategies of Bayesian
decision theory—namely, the strategy that an intelligent
88
investor or decision maker'is assumed to use when attempting
to maximize expected value.
CHAPTER V
BACKGROUND MATERIAL FOR ANALYSIS OF
CONGLOMERATE COMPANIES
The following equation denotes EpV, the expected present
value. This is an expression of the theory of value which
was defined as being appropriate for valuing the shares of
conglomerate companies under conditions of future uncer-
tainty.
n = m t = s Et(l + gtn)
fe
"pv ~ / / (1 + r)t E _ > > ,, . — * Prn
n = 1
Under conditions of future uncertainty, an investor
recognizes that on a given investment, the rate of return
actually experienced may not correspond—and probably will
not correspond—to the expected rate of return used in the
valuation. In the long-run, however, if an investor with
perfect expectations pays exactly the expected present value
for all securities, then the real rate of return experienced
will correspond to the expected rate of return.
The real rate of return experienced under conditions of
perfect expectations would not likely correspond to the
expected rate of return because the price paid for the
securities would tend to be different from expected present
89
90
value. In a decision-making context, it is possible to state
that an investor will attempt to buy shares at all times for
which the following ratio is the highest:
Expected Present Value
Price Per Share
If the ratio above were equal to one for all investments
made, then the long-run rate of return experienced under con-
ditions of perfect expectations would be equal to the
expected rate of return used in valuation. On the average,
if the ratio were more than one, then the actual rate of
return experienced would be more than the expected rate of
z'eturn used in valuation.
Perfect Expectations
For perfect expectations to exist, the investor must
have a perfect knowledge of the effects on gt of different
states of the world considered possible. Also, the proba-
bilities of different states of the world would have to be
known precisely.
It would, of course, be naive to assume that future
states of the world could be predicted precisely, which is
one of the reasons for the adoption of the expected present
value model. Also, it is impossible to attach exact proba-
bilities to different states of the world unless a
91
consistent process is generating future states of the world.
On the other hand, unless reasonable order does exist to the
generation of future states of the world and unless a
reasonable expectation of the effects of different states of
the world on g- can be attained, then valuation is not
possible. That is, for meaningful valuations to occur,
facts, relationships, and probabilities based on past
experience must be available which, though not necessarily
perfect, are valid enough so that decisions can be made which
are better than random ones. The hypotheses presented in
Chapter I and reproduced below are intended to provide the
basis for valuation of shares in conglomerate companies using
the expected present value model.
Hypotheses
(1) Conglomerate companies listed on the New York and
American stock exchanges acquiring the highest proportion of
publicly owned subsidiaries in the period 1961-1967, have
added negative value to them--where positive or negative
value added is measured relative to the weighted average of
the rates of return on book values of the resources employed
by the subsidiary companies prior to their acquisition.
Accordingly, negative value added is defined as the amount
by which the projected rate of return on investment exceeds
the rate actually experienced.
92
(2) The value added by conglomerate companies during
the period 1961-1967 has tended to decrease from year to
year as size and diversification have increased-—where
diversification is measured by the number of different
industries in which resources are employed, and size is
measured in terms of the book value of consolidated assets.
(The decrease in value added can be from negative values to
more-negative values or from positive values to less positive
values.)
(3) Based on a projection of past acquisition policies,
past weighted average growth trends of parent and subsidiary
companies, and the same percentage of value added by parent
company organizations—the current market prices of the
common stocks of conglomerate companies are too high if an
8 percent discounted rate of return is demanded.
(4) If the conglomerate companies surveyed are grouped
into (1) those adding the most value (relative to the other
companies) and (2) those adding the least value, then the
companies in each group will tend to have similar acquisition
strategies. Acquisition strategies for this purpose are
defined in terms of relative rates of return and growth rates
of subsidiaries and parent companies at the time of acqui-
sition.
Generally speaking, the belief underlying the presen-
tation of all of the hypotheses is that conglomerate companies
relying heavily on acquisitions have not added any additional
93
income potential to the subsidiaries that they acquire, but
rather have devoted time and resources controlled by manage-
ment to obtaining income through additional acquisitions
rather than through the release of synergy of the companies
acquired. That is, the hypotheses are intended to suggest
that the motivation for such companies has not been as much
to release synergy and create efficiencies as it has been to
make immediate additions to earnings per share.
The dependent variable of the first two hypotheses,
"value added," is only one factor contributing to income
experienced during a particular time period. Specifically,
"value added" is the efficiency of investment which can be
attributed to the initial parent company, synergy release,
and any unexpected increases in efficiency derived from
acquisition. The exact definition of "value added" is
elaborated more fully in the following sections of this
chapter. For the time being, "value added" can be viewed as
an efficiency variable which is largely controlled by the
parent company.
Hypothesis number one, if it is correct, indicates that
the effect on subsidiary growth potential of a company which
follows a strategy of acquiring a relatively large number of
publicly owned subsidiaries is, on the average, a negative
effect. Hypothesis number two states that the rate at which
value is added to subsidiary companies by a parent company
decreases as size and diversification increase.
94
Since companies which concentrate on acquisitions
generally are considered to depend upon a high P/E multiple
relative to the multiple of the companies acquired, it
follows that the subsidiary companies acquired may tend to
have relatively lower growth rates compared to the parent.
If such is the case, then it is likely that a slowing growth
or even a declining growth can occur as decreases in ope-
rating growth become large, relative to growth derived from
acquisitions.
The period 1961-1967 was chosen as the period for
testing the hypotheses for several reasons; mainly, however,
it was chosen because the broad economic circumstances in
which operating and acquisition policies have been conceived
and implemented have remained rather constant. By concen-
trating on this time period, the distortions in income which
can arise from changes in external conditions will hopefully
be minimized.
Problem Areas in Testing the Hypotheses
The hypotheses which were presented in the previous
section are to be tested by reference to past operating
performance during the 1960's. The conclusions derived from
the hypotheses are not useful, directly, in the valuation of
the shares of conglomerate companies because future and not
past operating performance is the source of current value.
However, in estimating future conditions, estimates must be
95
made relative to some starting point. The most fitting
reference is assumed to be the immediate past. In fact,
under conditions of future uncertainty, one state of the
world which can be considered possible is a continuation of
the immediate past.
In any event, before inferences can be made about the
future, as clear an understanding of the present as possible
is desirable. Unfortunately, the issues are clouded from an
economic, or public policy, point of view as well as from
the investor's point of view because of the problem areas
mentioned before and reiterated again below:
1. The lumping together of operating and acquisition
income.
2. The absence of established earnings trends under
various economic conditions.
3. The data regarding financial performance is generally
reported only on a consolidated basis.
4. Lack of uniformity of accounting methods between
conglomerate companies and before and after merger of firms
into a single conglomerate.
5. Lack of comparability of operating circumstances
before and after merger.
6. The presence of diversification of earnings due to
multi-industry orientation.
Recognition of all of the problem areas listed on this
page should be made in one form or another in the development
96
and application of a theory of value. Such recognition may
be manifested either in the theory of value itself, or in
the methods in which the theory is implemented, or even as
a limitation to the conclusions drawn. Hopefully, the limi-
tations imposed by the problem areas can be minimized.
In the second chapter a method of implementing the
present value theory was developed (i_.e. , the expected
present value theory), which gave recognition to problem
areas two and six. In this chapter, the effects of one and
three are examined. Also, the following topics are covered
as background material prior to testing the hypotheses:
(1) External conditions which have facilitated the
growth of conglomerate companies
(2) Methods of external expansion
(3) Pooling versus purchase accounting
(4) Some hypothetical transactions.
External Conditions Which Have Facilitated the Growth of Conflomerate Companies
It was stated in an earlier section that the time period
chosen for testing the hypotheses which were presented
earlier was chosen because it was characterized by rather
uniform external conditions. In the time period chosen,
1961-1967, the broad external economic conditions have been
characterized by
(1) General immunity of conglomerate companies to anti-
trust actions
97
(2) Historically high prices in the stock market
relative to "fundamental" values
(3) Sustained growth in GNP and other economic measures
of national output
(4) Historically high inflation rate, particularly
since 1965
(5) Historically high interest rates
Probably a few other factors could be added to the list
above; however, they are the main ones which can be used to
describe the period of gestation of most conglomerate com-
panies. It follows from this that conglomerate companies
are untested under other types of broad external economic
conditions. In fact, it is possible that conglomerates have
not yet been adequately tested by the conditions listed
above.
The general immunity of "pure" conglomerates from anti-
trust prosecution is one of the major factors which has
contributed to the growth of financially motivated companies.
That is, by concentrating on mergers which involve the combi-
nation of companies having similar functions and similar
needs (e.£., Procter & Gamble and Clorox), the antitrust
authorities have, to some degree, left the door open to pure
conglomerates. That is, the door has been left open to
mergers involving diverse firms in which the potential for
synergy release is the least. The truth of the dominance of
the conglomerate-type merger (where the term merger is used
98
generically to refer to all types of combinations) was indi-
cated by the statistics published by the FTC and .reporduced
in the first chapter.
If the possible motives for mergers can be classified
as financial, synergy release, and institutional,^ and if
synergy is ruled out by antitrust considerations (except
for financial and managerial synergies), it follows that
financial and institutional motives must account for the
growth of conglomerate mergers. If these two factors are
the main ones motivating the conglomerate merger surge, then
society's resources are probably not being employed more
efficiently in terms of pre- and post-acquisition status.
Stated differently, unless synergy is being released in spite
of these motives, then efficiencies are not being released
by conglomerate mergers.
It is not necessarily obvious, of course, that the
release of financial and managerial synergies, which are
theoretically possible for conglomerates, is an insignificant
factor in the conglomerate movement; however, it is also
quite possible that the main motives are financial or insti-
tutional .
Historically high prices in the stock market coupled
with investors' inability to place values on the shares of
conglomerate companies relative to long-run income potential
^See page 50 for a classification of motives underlying merger activity.
99
is also probably one of the- major factors which has contri-
buted to the growth of conglomerate mergers. Since shares
in conglomerate companies tend to be glamour-type stocks,
and for that reason are relatively more volatile than most
otners, periods of generally high prices are the periods in
which acquisitions can be implemented on the most favorable
terms. That is, in periods of high prices, fewer shares of
the parent company must be given up to match the market value
of the shares of the company being acquired. (The use of
higher-priced shares to acquire those of another company is
sometimes referred to as using "Chinese money.")
The historically high rate of inflation has probably
added to the rising stock market prices since it is a widely
held belief that stocks are a hedge against inflation. This
has undoubtedly stimulated demand for stocks on the part of
investors and contributed indirectly to the growth of con-
glomerate companies.
The sustained economic growth during the 1960's of
national product and income generally, is a facilitating
element in the merger surge because it helps to offset pos-
sible dilution in earnings per share that might have occurred
otherwise. Also, the growth in total revenue which can
reasonably be expected to coincide with general economic
growth would likely facilitate the attainment of higher
leverage positions.
100
If the broad economic circumstances mentioned earlier
should change, it is questionable whether or not conglome-
rates could sustain their current rate of mergers. Also, it
is questionable whether or not they could continue to support
their current leverage position. The effects on conglomerate
earnings would, of course, depend upon such factors as the
degree of actual diversification attained and how cyclical
the total earnings of the company would be, given the changes
in external conditions.
In the next chapter, some of the hazards awaiting the
unwary, financially motivated conglomerate and its stock-
holders are examined. These hazards are ones which could
arise, given the same external conditions in the future as
have prevailed in the 1960's. In the last chapter, the
effects which can arise from a change in external circum-
stances are suggested when different conditions are simulated.
Methods of External Expansion
Since this study is directed toward conglomerate com-
panies, one factor which must be dealt with is the methods
by which a company can become a conglomerate. The charac-
teristic which distinguishes conglomerate companies from
other companies which have followed the merger route to
growth is that conglomerates have acquired diverse firms
instead of those bearing significant horizontal, vertical, or
circular relationships. Accordingly, a firm can become a
101
conglomerate by acquiring diverse units, and this is almost
always done by external rather than internal means. The
methods of external expansion are the same for conglomerate
companies as for any other company. Namely,
(1) By merger
(2) By consolidation
(3) By purchase of assets
(4) By holding company
(5) By leasing
It is assumed that the distinction between the five
avenues of external expansion is reasonably clear so that a
detailed explanation of them is not necessary. Briefly, a
merger can be distinguished from a consolidation by virtue
of one company involved in the combination absorbing another
and retaining its own identity, while the company being
absorbed is dissolved and loses its identity. In the case
of a consolidation, two companies or more are brought to-
gether and a new one created with a new identity. Securities,
property, and cash can be exchanged in both types of trans-
actions .
Purchase of assets of one company by another is fre-
quently done to avoid getting approval of the stockholders
of the acquiring company. That is, in mergers and consoli-
dations, generally both companies' directors and stockholders
must approve the transaction ahead of time. In the case of
the purchase of the assets of one company by another, it is
102
not normally necessary for the acquiring firm to get approval
from its own stockholders.
Holding companies can be either holding-operating
operating companies or operating-holding companies, depending
upon the nature of the parent company. When the holding
company route to expansion is used, the subsidiaries acquired
or formed retain their own separate identity, and the stock
of these companies is treated as an investment by the parent
company. Special advantages accrue to this type of external
expansion, and it is probably the dominant form used by
conglomerates.
Leasing is, of course, one method of acquiring assets
externally. Title is not transferred through leasing, so
it may not really be classified as a method of external
expansion. In any case, leasing and purchase of assets as
methods of external expansion are not significant and can be
ignored for practical purposes as having no effect or
influence on the current merger movement.
Pooling Versus Purchase Accounting
It is possible for a parent company to acquire the
resources of another company by either merger, consolidation, •*>
Ernest W. Walker, Essentials of Financial Management (Englewood Cliffs, New Jersey, 1965), p. 92.
^Charles L. Prather, Financing Business Firms (Homewood, Illinois, 1961), p. 577.
103
or through a holding company and then simply merge the book
values of both companies together in its financial reporting.
To do this, however, the acquisition must be considered a
pooling of interest rather than a purchase. Pooling is
advantageous because, relative to the purchase method—which
requires recognition of premiums paid over book values,
calculations of income and rate of return on investment are
maximized. To illustrate the conditions necessary for
pooling, consider the following sequence of events.
Company M is a parent company conglomerate which openly
admits that it seeks out subsidiaries which will lead to an
immediate increase in earnings per share. Company M recog-
nizes the desirability of treating its acquisitions as
poolings rather than purchases; however, it must be able to
meet the criteria of the American Institute of Certified
Public Accountants regarding the conditions for poolings.
According to these guidelines a pooling is
. . . a business combination of two or more corpo-rations in which the holders of substantially all of the ownership interest (basically common stock) in the constituent corporations become the owners of a single corporation which owns the assets and businesses of the constituent corporations, either directly or through one or more subsidiaries, and in which certain other factors are present.4
The other factors which must be met if a combination is
to be considered a pooling include the following:
_^American Institute of Certified Public Accountants, "Business Combinations," Accounting Research and Terminology Bulletins (New York, 1961)," Bulletin No. 48, p. 22.
104
(1) The extent to which shares received by the several owners of one of the predecessor cor-porations are in proportion to their respective interests in such predecessor;
(2) If voting rights are materially altered;
(3) A plan or intention to retire a substantial part of the capital stock issued to the owners of one or more of the constituent corporations, or other substantial changes in ownership occurring before or planned to occur shortly after the combination;
(4) Abandonment or sale of a large part of the business of one or more of the constituents;
(5) The continuity of management; and
(6) Size of constituents.^
No one of the above factors would necessarily be
determinative and any one factor might have varying degrees
of significance in different cases. In short, the distinction
between when pooling accounting should be used and when
purchasing should be used is impossible to define precisely.
This situation has led, in the case of conglomerates, to
almost complete use of pooling. "In short, pooling has
become the 'in thing' and one must look far and wide before
one now finds an acquisition which is being accounted for as
a purchase."^
^Ibid., pp. 22-23.
^Abraham J. Briloff, "Distortions Arising from Pooling-of-interests Accounting," Financial Analysts Journal, XXIV (March-April, 1968), p. 73.
105
Some Hypothetical Transactions
As long as Company M does not need to worry too much
about meeting the requirements of pooling--since it apparently
can be justified rather easily—the main concern of Company
M is the efficiency of its acquisitions. That is, Company M
wants to acquire at all times in such a manner that its
acquisitions will have the most beneficial effect on earnings
per share.
A favorable candidate is selected, and Company M decides
to trade common stock for common stock with the company to
be acquired. Rather than getting a new authorization from
its shareholders, Company M decides to buy its own stock in
the open market (and artificaliy raise the price, thereby
improving its bargaining position with the company to be
acquired) and then trade the stock purchased for that of the
company to be acquired. Clearly, Company M is paying cash
for the company; yet, the final transaction involves an
exchange of stock, and it is therefore treated as a pooling
for accounting purposes.
Encouraged by the success of its first acquisition,
Company M decides to buy another through a cash tender offer.
By buying the shares of the company with cash, the need for
registration with the Securities & Exchange Commission and
other difficulties with the company's shareholders are
avoided. In fact, the cash tender offer can be arranged
so that, if the desired percentage of total ownership is
106
not acquired, the shares can be returned and the deal called
of f.
Company M is successful in its cash tender offer and
acquired controlling interest in the company. Company M is
now assured of voting control without effective resistance
from the remaining stockholders and the board of directors.
Also, Company M is in a position to exchange securities with
the subsidiary company on favorable terms and treat the
transaction as a pooling of interest because ownership is
continuous.
Instead of merging with the company just acquired,
Company M decides to operate as a holding company and borrow
money, using the stock of the company just acquired as
collateral. With the funds derived from the loan, Company M
purchases the common stock of an additional company. By
following this procedure, Company M knows that it will be
able to pyramid several levels of debt upon a single set of
productive assets. This is possible because debt at the
parent company level is backed primarily by its holdings of
stock in the subsidiaries. The subsidiaries, in turn, rely
on the real resources employed and their income potential as
collateral. Through pyramiding in this manner, the real
assets are supporting debt at the subsidiary level and again
at the parent level through income generated and transferred
to the parent company. If the subsidiaries can be encouraged
to become holding companies themselves, then it may be
107
possible to extend the overall leverage position even further.
That is, with three levels in the holding company, the real
assets at the lowest level could be used to support debt, at
least in part, at the two higher levels.
With three levels in the holding company, and debt
incurred at every level, Company M can be viewed, in terms
of the leverage position of a nonholding company, as a
company which has used its own securities twice as collateral
for debt. Obviously, this is a neat trick if it can be
carried out.
Being pushed for cash, Company M knows that investors
are rather naive, particularly with regard to conglomerate
companies, and decides that it would really be an astute move
to break up the subsidiary just acquired and form several
other corporations—letting the market try to value the new
shares. When the time is right, Company M sells off 10 to
15 percent of its holdings in the new companies and uses the
proceeds to help meet maturing obligations.
Company M is now pretty well conditioned to some of the
basic rules of the game; however, other considerations are
now necessary in its acquisition decisions. Because of
increasing needs for cash to support additional acquisitions
and the high leverage position, the company would be wise to
begin spending part of its time and resources looking for
companies to acquire which are of this nature:
(1) Under capitalized,
108
(2) Have low book values relative to liquidation value
of assets, and/or
(3) Have a favorable cash flow position.
Ideally, the companies acquired would add to earnings
per share and also possess one or more of the forementioned
characteristics. At some point, however, the company would
be willing to accept a slight dilution in earnings per share
in return for the added debt or cash which would be derived
from a company possessing the desired characteristics.
If the company acquired had low book values relative to
liquidation values, then it could be used to extend the over-
all leverage position of the firm. The same is true for a
firm which is undercapitalized—i_.e., additional debt could
be incurred at the subsidiary level and then magnified
through the holding company device. If the firm to be
acquired possessed a favorable cash flow position, it would
be attractive as a means for helping to meet maturing debt
obligations or to finance additional acquisitions.
In its acquisitions, Company M would prefer to give
convertible preferred stock or preference stock because
dilution of earnings per share of the common can be delayed
until additional companies have been acquired and additional
amounts of convertible preferred and preference stock have
been issued to take the place of those which have been con-
verted. Preference stock with an increasing conversion ratio
is preferable from Company M's point of view because such a
109
feature has a built-in incentive not to convert unless the
price of the company's stock declines.
While all of the moves open to Company M have not been
covered, a sufficient number have been illustrated to indicate
that growth through external means is, or can be, a rather
demanding process—demanding in that the company must not
only find acquisition candidates which add immediately to
earnings per share but also must arrange methods for their
financing, which is one of the determining factors in whether
or not earnings per share are increased. In fact, the
acquisition and financing processes could be so demanding
that, at least at the parent company level, little time would
be available for such small matters as releasing synergy—
particularly since, with a few exceptions, one of the major
characteristics of conglomerate companies is a relatively
small management staff at the parent company level.
While the method of financing and the organization used
for acquiring—.i.e., merger, consolidation, or holding
company—are important, if a company is to be successful as
a financially motivated conglomerate, it must, on the average,
make acquisitions which raise earnings per share. Of course,
as noted previously, some percentage of acquisitions can
occur because of their effect on cash flow or leverage
potential.
110
Acquisition Income
In the first chapter it was indicated that the income
per share of a conglomerate company could be broken down into
that which is derived from the parent company and that which
is attributable to the resources of the subsidiary companies.
The total income per share was denoted by the following:
k=t-l
V - E p < 1 + 9 p > t - Y + T e s U + g s )k ' z
T k=o
Y + tZ Y + tZ
parent subsidiaries
As the company described by the equation above moves
from one time period to another, income can be attributed to
two sources—operating income and acquisition income. By
definition, operating income is that income which is derived
from parent and subsidiaries which were acquired prior to
the current time period, and acquisition income is that
income which is attributable to subsidiaries acquired during
the current period. The distinction between the two sources
of income is made because acquisition income can be sued to
obscure the operating growth trend of a particular firm.
A firm can consolidate its reports with those of a
subsidiary acquired anytime prior to the closing of its own
books. Accordingly, it is possible, as demonstrated in the
first chapter, for a firm whose actual productivity is declining to show a gain in overall income simply by
acquiring on favorable terms. The use of acquisition income
Ill
to create a favorable impression was noted by Manuel F. Cohen
at a siminar at the University of Illinois:
There are also ways to increase a company's reported sales and earnings without improving performance—and here I speak of performance in its traditional sense. The easiest way, perhaps, is simply to add the sales and earnings of another company through merger or acquisition . . . And if, as has happened in some cases, the combined sales and earnings are compared with the unadjusted figures of the acquiring company for earlier periods, the increase in sales and earnings appears even more dramatic.
This accounts, in part at least, for the current rash of acquisition—hundreds, thousands of acquisitions~~and for the growth of the conglomerate company.7
For purposes of analysis, the equation
k=t-l Y t
Ep(i+gp)t-Y + Tes ( i + g s )k - z
Y + tZ k=o Y + tZ
is rather inconvenient. This equation assumes constant
forces are at work generating values for Yt. Such an
equation is applicable, perhaps, in attempting to project
future income levels; however, it does not give explicit
recognition to all of the leverage variables which play an
important role in the attainment of income per share through
acquisition.
^"Some Problems of Disclosure," The Journal of Accountancy, CXXV (May, 1968), p. 62, excerpted from an address by Manuel F. Cohen, Chairman, Securities and Exchange Commission, at the University of Illinois, Urbana, March 7, 1968.
112
The following equation is better for purposes of analysis
because it includes all of the variables, regardless of the
method of acquisition, which influences earnings per share
from one period to another. Also, the equation explicitly
recognizes return on total assets employed by the conglomerate
company rather than rate of return per share. Rate of
return on investment is viewed as a better variable than rate
of return per share for determining the "value added" by
conglomerates. The rate of return on total investment is
transferred to per-share earnings by recognition of leverage
variables. Accordingly,
E = (1-T) I (Ri ) (I)-(Rd)(D)] - (Rp) (P)-(Rm) (M) S " " " " ' I" » - - » .-.»•• il III • • LI I
s
In the above equation,
Es = Earnings per share during a time period
T = Tax rate
Rj = Rate of return on investment during the time period
I = Average investment level during the time period
Rd = Rate of return (average) on debt capital during the time period
D = Average amount of debt capital employed during the period
Rp = Rate of return (average) accruing to preferred stock
P = Average amount of number of pfd. shares outstanding
M = Dollar value of minority interest
113
= Rate of return per dollar to minority interest
S = Total number of shares of common stock outstanding
The equation for Es is better than the one expressing
Yt for purposes of analysis because it looks specifically at
the variables responsible for changes in earnings per share
during a time period. Of course, the equation for Es cannot
be applied to future time periods without being subject to
uncertainty, just as the equation for Y^. It is more
descriptive, however, when applied to past time periods.
Just as the equation expressing Y- expresses earnings
per share at some point in time "t," the equation for Es
could be used to express income per share (Es) as a function
of time if each of the independent variables were assumed to
be a function of time. In fact, each of the variables listed
above can be thought of as assuming different values per time
period (from year to year, for example). Es is the dependent
variable, and all of the other variables are independent
ones. Accordingly, if the value for one or more of the
independent variables changes from one time period to
another, there will be a corresponding change in Es. Thus,
if each of these variables is analyzed in an attempt to
explain past levels of income per share, a more meaningful
breakdown is possible than with the former equation.
Using the equation for Es, the following example is
presented to illustrate how a company can combine resources
114
which are declining in their individual growth rates (rate
of return on investment) and still show an increase in rate
of return from one time period to another. Also, the example
illustrates how the artificial growth can be magnified
through the leverage variables presented in the preceding
equation.
Acquisition Income of Company M
The financing and organizational policies of Company M
were discussed briefly in a preceding section; now, its
acquisition policy is illustrated. Company M has the
following characteristics:
Total Investment $200,000 Total Debt $100,000 Average Interest Rate on Debt 5% Annual Interest Expense $ 5,000 Total Equity $100,000
Retained Earnings $ 50,000 Common Stock (20,000 shares outstanding) Preferred Stock ($5 annual dividend)
100 shares outstanding Tax Bracket 50% Rate of Return on Investment
(before interest and taxes) 10% Price Earnings Multiple on Common Stock .20 to 1 Growth Rate of Rate of Return on Investment zero Annual Gross Income . $20,000 Income after Interest Expense $ 15,000 Income after Taxes and Preferred Dividends $ 7,000 Income per Share 35<? Price per Share $ 7.00
As indicated in the preceding financial information,
Company M has a zero growth rate in its rate of return on
investment. If rate of return on investment is viewed as an
index of efficiency and profitability, then it is possible
115
to state that Company M is remaining the same from one time
period to another in terms of these characteristics. The
reason that the parent company, Company M, is able to command
as high a price-earnings multiple as it does is that it has
been able to give the investment community the impression
that it is a growth company by making acquisitions which give
the impression that total rate of return on investment is
increasing.
To illustrate how Company M is able to give the im-
pression of being a growth company through its acquisitions,
let it be assumed that the company acquires subsidiary
companies with the following characteristics in each of four
subsequent time periods:
Total Investment $100,000 Total Debt $ 25,000 Average Interest Rate 5% Average Interest Expense $ 1,250 Total Equity . $ 75,000
Retained Earnings . . . $ 45,000 Common Stock (10,000 shares outstanding) Preferred Stock (none)
Tax Bracket 50% Rate of Return on Investment
(before interest and taxes) 15% Price-Earnings Multiple of Common Shares 10 to 1 Growth Rate of Rate of Return on Investment
. . . . a minus one percentage point per year Gross Income in Year of Acquisition . . . $ 15,000 Income after Interest Expense $ 13,750 Income after Taxes $ 6,875 Income per Share $ .6875 Price per Share $ 6.875
Under the assumption that the "M" Company acquires a
subsidiary with the above characteristics in each of four
subsequent time periods, it is clear that the rate of return
116
on total investment employed is d e c l i n i n g — , the parent
has no growth, and each subsidiary has a minus one percentage
point per year.
While it is quite obvious from the preceding example
that the overall growth rate is declining, notice the growth
trend that the parent company is able to report because of
its acquisitions:
Period Number Zero (Parent Company Alone) Rj = 10%
Period Number One (Parent plus one Subs.)
RT = (200,000) (10%) + (100,000) (15%)
3 0 0 , 0 0 0
RJ = 1 1 . 6 7 %
Period Number Two RL = (200,000) ( 1 0 % ) + ( 1 0 0 , 0 0 0 ) (14%) + (100,000) (15%)
4 0 0 , 0 0 0
RJ = 1 2 . 2 5 %
Period Number Three
RJ - Same Numerator as Above + ( 1 0 0 , 0 0 0 ) (.13%) 5 0 0 , 0 0 0
RJ = 1 2 . 4 0 %
Period Number E'our
R t = Same Numerator as Period Three + (100,000) (12%) 600,000
RJ = 1 2 . 3 4 %
The parent company can continue to sustain increases in
Rj only so long as it can acquire companies whose Rj's
contribute more to current Rx than the negative growth rate
117
takes away. In spite of the eventual decline in Rj, Company
M was able to give the impression over the entire acquisition
period of having an increasing trend in Rj.
The preceding example suggests that a company can give
the impression of being a growth company when it is actually
declining in terms of its rate of return on investment. This
illusion is not limited to overall rate of return but is
transferred to earnings per share (and very likely magnified
by changes in the leverage positions of parent and subsidi-
aries acquired). As noted in the first chapter, for income
per share to be added through acquisition, the following
must be true:
T e s must be greater than Ep
in
Z(Tes - Ep) g Y + z
Where
Ep = Current earnings per share of parent company
Tes= Earnings per share of the resources of the acquired company after shares of the parent company are traded for those of the acquired company
Z = The number of shares traded of the parent company for those of the acquired company
Y = The number of shares of the parent company before acquisition.
The same relationship expressed in the previous equation can
be defined more generally in terms of changes in the vari-
ables of the following equation:
118
E ^ (1-T) I (Rj) (I)-(Rd) (D)]-(Rp) (P) s g ™
Where, as before
Es = Earnings per share
T = Tax rate
Rj = Rate of return on investment during the time period
I = Average investment level during the time period
R(j = Rate of return (average) on debt capital during the period
D = Average amount of debt capital employed during the period
Rp = Rate of return (average) accruing to preferred stock
P = Average amount of (number of) pfd. shares outstanding
S = Number of shares of common outstanding at the end of periods
If any one of the above variables changes due to acquisition
of a subsidiary, then there will be a corresponding change
in earnings per share—this is true also of changes arising
from one period to another from internal operations. The
following section presents the effects on earnings per share
attributable to Company M's acquisitions.
Effects on Earning per Share of the Preceding Acquisitions
Earnings per share (assuming a 50% tax rate) in period
number zero are
E _ (.50) [(.10) (200,000) - (.05) (100,000)] - ($5) (100) s " 20,000
119
Es = 35 cents
Earnings per share in period number one: (Assuming that the
company acquired is paid for with 5% bonds and $5 preferred
stock, and also assuming that the company maintains its one-
to-one debt-equity ratio and the same ratio of preferred
stock to total investment.) Note that the acquisition terms
are subject to negotiation and, so long as the parent company
can acquire companies whose total earnings exceed fixed
charges assumed in the acquisitions, then some contribution
is made to equity holder's income. In the case at hand,
(.50) [(.1167) (300,000)-(.05) (150,000)]-($5) (150)
Es = 20,000
Es = 65 cents
Earnings per share in period number two: (This year's
earnings are different from the previous period's income
because of the acquisition of an additional subsidiary and
also because of the decline in the rate of return on the
subsidiary acquired in period number two.) To isolate the
effect on income per share from acquisition alone, it would
be necessary to subtract the earnings per share that would
have occurred without acquisition from that which occurred
with acquisition. Since this is not done, the income per
share calculated below is the cumulative effect of changes
in the rate of return on investment from all preceding
acquisitions.
120
(.50) I (.1225) (400,000)-(.05) (200 , 000)]- ($5) (200) Es = 20,000
Es = 93 cents
Earnings per share in period number three:
(.50) I (.124) (500,000)-(105) (250,000)]- ($5) (250) Es = " 20,000
Eg = $1.18
Earnings per share in period number four:
(150) [(.1234) (600,000)- (105) (300,000)]-($5) (300) Es = 20,000
E s = $1.40
Summary of Operating and Acquisition Performance of Company M
The examples presented in Figures 14 through 20 should
illustrate clearly that not only is it possible for increases
in rates of return on investment to be passed on to common
stockholders, but, also, such increases can be greatly mag-
nified in terms of earnings per share. Even more significant
is a conglomerate's ability to give such an impression of
growth when, in reality, the productivity of resources is
declining. The secret lies in the calculation of growth
rates for conglomerate companies relative to the past period
rate of return for the conglomerate and not relative to the
past rate of return for the resources employed--_i. e.,
traditional calculations are not geared for companies
which acquire most of their productive resources secondhand.
Ri
12%
11%
10%
0 1 2 3 4
Company M.
Total Investment
600,000
500,000
400,000
300,000
200,000
121
RI Time
10 % 0
11.67% 1
12.25% 2
12.40% 3
12.34% 4
on investment for
Total Investment
Time
200,000 0
300,000 1
400,000 2
500,000 3
600,000 4
0 1 2 3 4
Fig. 15--Trend in total investment for Company M
122
(I)(R z)
80,000
70,000
60,000
50,000
40,000
30,000
20,000
N
0 1 2 3 4
Time
Fixed Charges (debt and taxes)
45,000
40,000
35,000
30,000
25,000
20,000
15,000
10,000 0 1 2 3 4
-Time
Total Time Revenue
20,000 0
35,000 1
49,000 2
62,000 3
74,000 4
for Company M
Fixed Time Charges
12,500 0
21,255 1
29,500 2
37,250 3
44,520 4
for Company M
123
Net Contribution to Equity
(1-T){(Rx)(I)-(Rd)(D)J
30, 000
25, 000
20, 000
15, 000
10, 000
5, 000
Net Contribution Time
7,500 0
13,745 1
19,500 2
24,750 3
29,480 4 0 1 2 3 4
Time
Fig. 18-~Trend in net contribution to equity for Com-pany M.
Net Contribution to Equity Minus Preferred Income
(1-T) [ (Rj) (I) - (Rd) (D) ] - (Rp) (P)
30,000
25,000
20,000
15,000
10,000
5,000 0 1 2 3 4
• -T ime
Net Contribution Time to Equity Minus Preferred Income
7,000 0
12,995 1
18,500 2
23,500 3
27,980 4
Fig. 19 Trend in net contribution to equity minus pre-ferred income for Company M.
124
Earnings per Share
(1-T) [(Rj) (I)-(Rd) (D)]-(Rp) (P)
$1.40
$1.20
$1.00
$ .80
$ .60
$ .40
$ .20
Earnings per Share
Time
$ .35 0
$ .65 1
$ .93 2
$1.18 3
$1.40 4
0 1 2 3 4
Time
Fig. 20--Trend in earnings per share for Company M
Effects on Market Price of Common Stock
One of the initial assumptions was that the investment
community viewed the "M" Company as a growth company (because
of its rising Rj), and, accordingly, the community was willing
to pay a price-earnings multiple of twenty to one. Under
these assumptions, the price per share in period one would
be $7.00; in period two it would be $13.00; in period three
it would be $18.60; in period four, $23.40; and in the last
period, $28.00. Or, for the entire period, the market price
of the stock would have appreciated by more than 300 percent,
while the company's total resources employed continued to
become less productive.
125
How Long Can Acquisition Be Used for Growth Purposes?
Unfortunately, the growth rate of the conglomerate is
the investment-weighted average growth rate of all groups of
resources employed; therefore, once a company becomes com-
mitted to a growth policy from acquisition alone, larger or
more frequent acquisitions become imperative if the past
growth rate is to be maintained. If, as in the case of
Company M, a company becomes committed to a policy of
acquiring companies with negative growth rates for their
immediate effect on Rj—progressively larger acquisitions
must be made to offset the effects of the overall negative
trend, and at some point earnings will become negative.
CHAPTER VI
GROWTH TRENDS UNDER CONDITIONS OF
SUSTAINED ACQUISITIONS
For discussion purposes, the effects of a parent
company's acquisitions in a series of time periods can be
viewed as a repetition of the effects of the single acqui-
sitions which were presented in Chapter I. Recall that in
terms of relative values for T e s, Ep, gp, and gs, there are
six possible combinations for describing single acquisitions:
(1) T e s>Ep and gs> gp
(2) T e s > Ep and g s = gp
(3) T es> Ep and g s < gp
(4) Tes< Ep and gs>gp
(5) T es< Ep and gs = gp
(6) T e s < Ep and g s < g p
Of the six combinations listed above, three are not
feasible as models for the long-run strategy of firms
attempting to add earnings per share through acquisitions.
They are not feasible because in these three particular
instances, where Tes is less than Ep, earnings per share are
diluted rather than increased by acquisition.
The three remaining strategies (1, 2, and 3 above)
make immediate additions to earnings per share. Of the three
126
127
strategies, number one is the most desirable, number two is
the next most desirable, and number three is the least
desirable. This is the order of preference from the parent
company's point of view because in situation one both imme-
diate earnings per share are increased and the operating
growth trend of the total company is improved. In situation
two an immediate addition in earnings per share is made, but
the operating growth rate remains the same for the company
before and after acquisition. In situation three an imme-
diate increase in earnings per share is made, but the long-run
operating growth trend is diminished.
Unfortunately, since the relative operating growth rates
of the companies determine, or should logically determine,
the terms on which a parent company can acquire a particular
subsidiary, the combinations most easy to consummate with a
positive immediate effect on earnings per share are the ones
having the least favorable effect on the long-run operating
growth trend. Accordingly, there would be a tendency for
acquisition candidates in terms of positive Eg's to be ranked
(3), (2), and (1)—in that order. That is, the most favorable
condidates for merger in terms of both immediate income
effect and effect on operating growth trend are the ones
hardest to acquire on favorable terms if the overriding
factor in the exchange of ownership is the relative growth
rates of the parent and subsidiary.
128
In terms of graphs, the three positive acquisition
models are presented in Figure 21:
Yt
Er
(1) Tes > Ep gs> go
Time
(2) T e s> Ep gs - gp
Yt
Ec
Time.
Yt
Eg {
(3) T e s > E p g s < g p
-Silbsidia^
"Time
Fig. 21—Trends reflecting positive acquisition income from single acquisitions.
The equation for the resultant follows:
129
y = Ep (1+gp)t,Y + Tes (l+gs^'Z t Y + Z Y + Z
If the long-run trend of income per share of a conglome-
rate is viewed as a series of resultants—one building on
another, then the growth trend over time might look similar
to the following for each of the three feasible acquisition
strategies: (where Tes and gs remain constant for each sub-
sidiary acquired)
(1) T e s > Ep and g s>gp
Time-
(2) T e s> Ep and gs = gp
•Time-
Tes > EP a n d 5s<gp JP
_Txme.
Fig. 22--Long--run trends resulting from sustained acqui-sitions which add immediately to earnings per share.
130
The equation for the zig-zag line in each of the three
cases is
Ep(l+gp)t-Y |rt Tes(l+gs)k'Z
Yt = _ _ _ + N Y + tZ k^O Y + tZ
and
Egt
where
(Tes ~ Yt*)Z Y + tZ
E gt . . . . denotes immediate increase in earnings per share as a function of time
*t' . EP ( 1 + gP } t' Y
+ Tes (1 + gs)^'z
Y + tZ k=l Y + tZ
Figure 23 magnifies case number one of Figure 22. In the
diagram presented in Figure 23, the upward operating trend
of both parent and subsidiaries acquired manifested in an
increasing Yt* causes each successive Egt to be smaller than
the previous one. This means that the continued acquisition
of the same subsidiary on the same terms per time period will
have a decreasing absolute effect on total income per share.
Also, in percentage terms, the effect of the subsidiaries'
operating growth trends will have a greater effect in earlier
time periods than in later ones because the total operating
growth trend is the investment-weighted average of parent and
subsidiaries, and the size needed to have the same effect
becomes progressively larger. This means that the overall
131
S >E P <?s>gp
*t
% 3
Eg2
Egl
0 1 2 3 4 5 6
Time
Fig. 2 3—Long-run trend in earnings per share when acquisition income is positive and gs is greater than gp.
growth trend attributable to acquisition income and operating
income is highest in earlier periods. Accordingly, if one
attempted to make a future projection of earnings based upon
income generated during early periods, the projection would
not be realistic when compared to that which would transpire
under the conditions assumed to exist.
In order to sustain the same growth trend through
acquisition in later periods, it would be necessary for the
parent to increase the size or rate of acquisition or to
acquire on more favorable terms than in earlier periods. In
any case, the long-run trend will approach its operating
trend due to limits which exist on the ability of a firm to
continually increase in a favorable manner the size, rate, or
terms of acquisition.
132
In case number two, the same generalizations can be
drawn as were drawn about case number one; however, Eg-j- does
not decline in absolute terms as rapidly as it did in case
number one. Eg^ does not decline as rapidly because the
operating growth trends of the subsidiaries acquired are less,
relative to those in case number one. In case number three,
the influence of acquisition income is more pronounced and
endures longer because Y^* is not increasing as rapidly in
this instance as it was in cases one and two. It is not
increasing as rapidly because of the relatively lower growth
trend of the subsidiaries being acquired.
In all three cases, if the overall growth trend of
parent and subsidiary combined is negative, then, as illus-
trated in Chapter I, acquisition income can be used to create
the impression of an overall growth trend only so long as
the total decrease from operating earnings is offset by
increases from acquisition. Beyond some point, however,
earnings per share will begin to decline absolutely.
In all three cases, when the operating trend is positive,
the effects of acquisition are more pronounced in early
periods than in later ones; this means that earnings pro-
jections of past experience are biased in a positive manner,
but to varying degrees—depending upon operating growth
trends of subsidiaries acquired. The positive bias in early
periods means that any projections of earnings will be
inherently high, relative to the operating trend upon which
133
acquisition income is added. This situation can be viewed
in graphic form in Figure 24.
Projection on basis of past trend
Time
Gap
Projected operating trend
Fig."24—Gap between operating trend and total income trend.
It is the gap between operating trend and total per-
share earnings trend which can allow a conglomerate company
to acquire subsidiaries which make a positive immediate con-
tribution to earnings and also contribute to the overall
operating trend. That is, in its bargaining with a sub-
sidiary, a parent company can point to the earnings trend
which includes both past operating and past acquisition
income and compare it to the operating trend of the sub-
sidiary.
It is possible, and even probable, that in many instances
the combined trend of acquisition and operating income of
the parent is greater than the operating trend of the parent
while the operating trend of the parent is less than that of
the subsidiary. Under such conditions, the parent company
134
is in an excellent position to add to both components of its
income per share. Of course, as the preceding examples
illustrated, a company is able to sustain its total trend in
earnings only so long as it is able to increase in a favorable
manner the size, rate, or terms of acquisition; unless this
is done, the gap between total earnings per share and
operating earnings will begin to close. In the interim, the
unwary subsidiary and its stockholders can be deceived.
The source of the deception lies in the inapplicability
of the traditional concept of growth to financially oriented
conglomerates. That is, growth through acquisition is a
different form of growth than that which occurs through
internal means. Normally, growth through internal expansion
carries with it the implication that the firm is experiencing
a strong demand for its products because of some special
attribute of the product or the production process from which
the product is derived. Furthermore, such demand is normally
expected to endure for an extended time period, perhaps
following a geometric pattern over time. On the other hand,
growth through external means can occur when actual demand
and actual operating profitability are declining; also, such
growth can be terminated by such factors as unfavorable
swings in the stock market or changes in the attitude of the
antitrust authorities. Clearly, a distinction should be
made between the two types of growth.
135
Difficulties Encountered in the Valuation of Shares in Conglomerates
As noted in the previous section, anyone attempting to
value shares in conglomerate companies must reconcile
numerous variables. The presence of numerous variables cause
the following circumstances to arise:
(1) At any point in time, the earnings per share are
a product of both operating income and acquisition income.
Furthermore, if it is assumed that the subsidiaries which are
acquired, on the average, are ones which add immediately to
earnings per share, then the trend implied by the past record
will be inherently higher than the operating trend.
(2) There is a continually changing operating growth
trend in earnings per share if the resources acquired have
different growth rates from those of the parent and other
subsidiaries.
(3) Future income levels depend on size, rate, and
terms of acquisition, as well as on operating income.
(4) Changes in the overall leverage position influence
earnings per share.
The gap between operating trend and total income trend
is graphically illustrated in Figure 25.
Sources of Income
If an observer is at time period four of Figure 25, then
the past trend in earnings per share which is observed is the
combined trend from both acquisitions and operations. In
-Future
Time
136
Projection on basis of past trend
Gap
Projected operating trend
Fig. 25—Gap between operating trend and total income trend.
order for a projection of the past trend into the future to
be realistic, the same percentage increases in earnings per
share must arise collectively from the following sources of
income:
(1) Size of acquisitions,
(2) Profitability of acquisitions,
(3) Rate of acquisitions,
(4) Synergy release,
(5) Internal expansion of income, and
(6) Changes in the leverage position.
That is, in order for a conglomerate company to sustain its
past growth rate or keep the gap between operating earnings
and acquisition earnings from closing, it must increase from
period to period one or more of the preceding variables,
relative to the prior period.
137
Under current conditions, it is impossible to determine
precisely which of the sources of income are the major ones
responsible for the dramatic growth in earnings per share of
many conglomerates. Conglomerate companies' spokesmen aver
that it is sources four and five which account for a con-
siderable part of growth, and, unless these are sources of
growth, it is questionable whether or not such growth can be
sustained.
The hypotheses presented in the preceding chapter were
derived to test the proposition that, contrary to what is
being suggested by the spokesmen for conglomerates, the main
sources of growth to conglomerate companies are not four and
five, but some combination of the others. Pursuant to this
objective, the equation for below cannot be used effec-
tively to isolate the changes in income from all of the fore-
mentioned variables.
Y - EP ( 1 + gP ) t' Y + ^ Tes(l+gs)
k'Z Y + tZ K^o Y + tZ
Definitions:
(1) Real variables—Rj, I, gRj, and gl
(2) Leverage variables—T, R^, D, P, Rp, M, Rm, and S
(Rm refers to rate of return paid to minority interest and
M refers to minority interest.)
For purposes of analysis, the following equation for Y .
is more appropriate than the one above. This equation is
more appropriate because it gives recognition to both the
138
real variables and the leverage variables which determine
U-T) [(RIp) (Ip)-(Rdp) (Dpj) - (Rpp) (Pp)-(Rmp) (Mp)
Sp
plus
^ a-n ^=0 Cl-T) (RIa) Cla)-(Rda) (°a) - (Rpa^ pa)-(% a) (
Ma) a=o S a
In these latter expressions, the subscript "p" denotes
variables which are descriptive of the parent company, and
the subscript "a" denotes the values for each of "n" sub-
sidiaries that are acquired. To be used to project future
values for Y^, each of the independent variables in the
equation must be endowed with orderly values over time.
The second of the two terms in the equation expresses
the net contribution to earnings per share by the subsidi-
aries acquired, where such subsidiaries are assumed to be
financed in a specific and identifiable manner. In reality,
it is not always possible to determine how individual
acquisitions are financed—particularly if cash or some
combination of cash and other securities is given up during
acquisition. Also, it is difficult or impossible to distin-
guish between sources of financing internal and external
expansion. Accordingly, the following equation is a restate-
ment of the preceding on in terms of combined real variables
and combined leverage variables for both parent and sub-
sidiaries :
139
a~n ' (1-T) 1(1 ) (Rj ) + (Ia) (Ria)-(Rd) (D)]-(Rp) (P)-(Rm) (M)
Yt ~ 1_ aT°. s
Again, as before, in order to be a predictive model, each of
the preceding variables must be considered to follow some
pattern of behavior over time.
Real Variables Expressed as Functions of Time
If, for the time being, attention is directed only to
the real variables in the preceding equation, and if these
variables are expressed as functions of time, then gross
income is as follows:
a=t (it) CRt> =i pd+gip)t-Ri pd+gRip)i a(i+gia)
t _ a , Ri a<1 +9 Ri a>
t _ a
cl—O
Gross Contribution to Gross Contribution to Gross Income Income by Parent Income by Sudsidiaries
In this equation, It is total investment, Rt is rate of
return on total investment, and t is time. The equation is
defined so that all of the acquisitions of a single time
period are treated as one acquisition, and the related real
variables are the average ones for the period. From one
period to the next, however, recognition of different values
for the real variables associated with acquisitions are
recognized. For example, letting a different subscript
denote a different period's acquisitions, when t = 0,
Contribution to Gross Income by Subsidiaries is— (IQ)(R0)
140
When t = 1, the Contribution to Gross Income is— (I0) (l+gl0)^ (Riq)(l+gRi0)
plus
Ul) (Rl)
When t = 2, the Contribution to Gross Income is— (I0)(l+gl0)2 (Ri0)(l+gRj0)^ plus
(11)(l+gl1)l (R1±) (l+gRI;L)l plus
(12)(r2>
and so on.
Solving the preceding equation for R -, it becomes
Rt ipCl+gip)t-Rip(i+gRip)t aptia(l+gia)
t a*Ria(l+gRia)t~a
lt a=o xt
Term A Term B
If a parent company with the real variables denoted in
the preceding equation acquires subsidiaries per time period
such that Rja is greater than the Rt that would otherwise
have occurred, then the pattern of Rt generated by a series
of acquisitions might look similar to the following:
I t /v
R.J. Based on Past
Operating Rt
Time
Fig. 26—Divergence between past trend and operating trend.
141
In other words, acquisitions can create a gap between pro-
jected trend (trend from both operating and acquisition
sources) and the operating trend—just as in the case of
earnings per share.
Value Added
If the real variables contained in the previous
equations are redefined as the ones which would have occurred
if the parent and subsidiaries had not been combined, then
Changes Occurring R-f- = Term A + Term B + Because of
Combination
If changes in rate of return on investment occurring because
of combination are considered to be the result of synergy
release, which can be positive or negative, then the latter
term in the equation above can be considered a measure of
the amount of synergy released in the combination.
As a practical matter, due to the consolidation of
income and the physical combination of parent and subsidiary
companies, the separate operating trends of parent and sub-
sidiaries are obscured; therefore, Term A and Term B cannot
be determined precisely after combination. Because of this
limitation, the following modifications are introduced: (1) g!a and gRl^ are allowed to be approximated for
each subsidiary by the trend occurring prior to acquisition
(2) Term A is replaced by— (ID)(RT ) (denoted by A') ' p
XP +g_fc da) (i+gia)t-a
a=o
142
(3) Term B is replaced by— (denoted by B')
a=t y ~ ia(i+gia)
t"a-Ria(i+g%a)t"a
a=o a=t
Ip + Iad+gla)^3
a=o
(4) "k" is the multiple needed to equate Rt actually experienced with Term A' plus Term B*
Implementing the modifications listed above/ the
expression for R-j- becomes
R_ = k [Term A' + Term B']
or
R4
a=t < y ( R i P '
a=t "P "'"a (l+g^-a) a
a=o
+ a=o iad+gia)^'-^ (1+gRj )
a xa t-a
a=t Ip + H Ia ( 1 +9 Ia>
t-a
a=o
For a particular conglomerate company under scrutiny,
it is possible that all of the variables in the preceding
equation could be known except k. Accordingly, for purposes
of analysis, k could be treated as the dependent variable.
If k were equal to or less than one, then the net effect on
Rt of the following factors was zero or negative during the
time period under consideration:
(1) Divergences in gla and gRia from the trend which
existed immediately prior to acquisition;
(2) Synergy release; and
(3) Internal expansion of the parent company.
143
The main significance of "k" is that if it can be
determined that during a time period a conglomerate company
experienced a "k" equal to or less than one while, at the
same time, its "R -" increased, the only source of growth was
through acquisition. Stated differently, if "k" is equal to
or less than one, the parent company added no value to the
subsidiaries that it acquired during the period—in spite of
the fact that "R^" increased for the conglomerate. A con-
glomerate company with a "k" equal to or less than one which
has an increasing "R^" is deriving the increases by virtue
of its acquiring subsidiaries which have progressively
greater effects on "Rt" and not because it is adding value
to them.
Value Added by Company M
Company M, whose acquisition policies were examined in
the preceding chapter, is an example of a company which added
no value to its subsidiaries but was able to show an
increasing Rt. For Company M, k equals 1, and
(I M R . ) 2Ila(l+gIa)t-a-R (i+gR )t-a R+. = — - e + a=.o t
a=t a=t XP + 2 _ I a ( 1 + g I a ) t _ a Ip + 2 1 I a(l+gl a^
t _ a
a=o a=o
In other words, Company M is a parent company which has no
internal expansion influencing Rt, no synergy release, and
no change in R t is derived from its subsidiaries other than
that which is projected. In spite of this, Company M was
144
able to show an increasing Rt while simultaneously acquiring
companies with negative growth trends.
In the case of Company M, all of the increases in R-{-
which were derived can be attributed to the growth rates
implicit in the parent and subsidiaries prior to acquisition
and the relative sizes of acquisitions. In other words, the
parent company did not add any operating value to the sub-
sidiaries, in spite of the fact that the subsidiaries alone
would have reported a declining growth rate.
In the following chapter, a selected group of conglomerate
companies are examined, and an attempt is made to estimate
the "k" associated with each of them during a time period.
Hypothesis number one, which was presented in an earlier
chapter, suggests that the "k's" will, on the average, be
negative. Hypothesis number two suggests that the
growth in "k" from year to year is negative as size and
diversification increase.
Due to problem areas surrounding analysis of conglomerate
companies, several additional modifications are made to
"value added," defined in this chapter. The nature of these
problems and changes imposed by them are examined in the next
chapter prior to testing the hypotheses.
CHAPTER VII
RECOGNITION OF PROBLEM AREAS AND
TESTING OF THE HYPOTHESES
In an earlier chapter the following problem areas were
mentioned as being factors inhibiting analysis of conglom-
erate companies:
(1) The lumping together of operating and acquisition
income.
(2) The absence of established earnings trends under
the various economic conditions,
(3) The data regarding financial performance is generally
reported only on a consolidated basis.
(4) The lack of uniformity of accounting methods
between conglomerate companies and before and after merger
of firms into a single conglomerate.
(5) Lack of comparability of operating circumstances
before and after merger.
(6) The presence of diversification of earnings due to
multi-industry orientation.
If conglomerate companies are examined in terms of the
performance of subsidiaries prior to acquisition relative to
postacquisition, at least some of the difficulties created
by problem areas one and three can be eliminated. That is,
145
146
if conclusions about conglomerates are based on a concept of
"value added" after acquisition, it does not matter that
operating income and acquisition income are lumped together,
and it does not matter that data is provided only on a con-
solidated basis, so long as adequate data can be determined
for companies prior to their acquisition and for the combined
parent and subsidiaries in subsequent time periods.
In order to implement the expected present value model,
it is necessary to have financial data provided on the basis
of groups of resources whose operating income potential
responds similarly to specific changes in external con-
ditions. In so far as this is true, the consolidation of
income in financial reports represents a significant limi-
tation. Since more detailed data is provided on the 10-K
reports sent to the Securities and Exchange Commission, it is
possible that reference to these reports can yield better
results. However, the 10-K reports fall short of actual
requirements.
Problem area five, the lack of comparability of operating
circumstances before and after merger, makes it difficult
or impossible to rely on accounting data to draw conclusions
about efficiency of resource utilization before and after
merger. For example, the parent company's bargaining
position may enable the newly acquired subsidiary to maintain
a stronger position relative to outside parties, such as
suppliers, which will be reflected in increased profitability.
147
Also, tha pricing policy that might be imposed by the parent
company on its subsidiary could have an effect on income
streams generated after acquisition. For reasons such as
these, reliance on accounting data to draw conclusions about
efficiency in "real" terms is questionable. Accordingly, to
partially avoid this limitation, the viewpoint of an investor
has been assumed throughout this discussion, and when
reference is made to changes in efficiency, it is a change
in income potential only—with no implication intended
regarding efficiency of resource utilization.
Problem area four, the lack of uniformity of accounting
methods, imposes probably the greatest constraint of all on
an individual's ability to draw valid conclusions about con-
glomerate companies. Since the major independent variables
being analyzed are financial ones, and since some degree of
variance can exist in accounting valuations of transactions,
it is customary for analysts and others engaged in financial
investigations to make certain adjustments in published
financial statements. However, in the case of conglomerate
companies, there is more than one source of accounting
difficulties. Not only are traditional accounting procedures
a limiting factor, but, also, there is no general agreement
about what information should be furnished to those interested
in conglomerate companies.
Studies have been conducted and others are in progress
regarding the form and procedures which should be followed
148
.in reporting to investors and potential investors who are
interested in conglomerate companies. Probably the most
noteworthy and most recent of such studies was the one
conducted by Robert K. Mautz for the Financial Executives
Research Foundation. This study was largely devoted to a
survey of various individuals' opinions about what changes in
reporting methods should be made to make financial reporting
for conglomerate companies more meaningful. In a question-
naire sent to a large number of analysts, the following
objectives and criteria for analysis were recommended:
The investors' questionnaires indicate that analysts rate "maximum return in the long run from a combination of dividends and capital appreciation" as the most important objectives for those responding to the questionnaire.
The most important company characteristic in attaining these objectives are indicated as growth potential, managerial ability, and profitability, in that order.
The return on common stock equity, return on total assets and ratio of net income to sales were rated as the most useful indicators of profitability.
The preferred indicators of growth potential are growth of major markets, rate of growth in earnings per share, and research and development expenditures.
Managerial ability is best indicated by the growth of the company, the return on common stock equity, and the personal reputation of key per-sonnel . 1
•'-Robert K. Mautz, "Findings and Recommendations Released on Financial Executives Research Study," Financial Executive, XXXVI (February, 1968), p. 53.
149
In the preceding recommendations, most of the criteria
and characteristics suggested for measuring the ability of
conglomerate companies to attain the goals of "maximum return"
were characteristics and criteria descriptive of companies
already in a position of satisfying the goals of its stock-
holders. That is, the criteria and characteristics recom-
mended were not decision criteria, but identification
criteria. For example, it does not require a chartered
analyst to determine that profitability is normally synono-
mous with a favorable rate of return. Such criteria amount
to little more than an exercise in semantics, in which one
effect is restated in terms of another. The chain of circular
reasoning was completed when, in the last section of the
recommendations, return on common stock equity was deemed a
good indicator of managerial ability--that is, suggesting
that the effect (favorable return) is an indicator of one of
the causes (managerial ability). Such statements are quite
true, but they do not deliver much information for investors'
decision-making purposes.
Probably one of the major reasons for the unrewarding
results of the forementioned survey is that the persons being
surveyed did not have any preconceived notion regarding what
use they would make of any information after it was provided.
In order to know what information was needed, the persons
being surveyed would need to know in advance what use would
be made of it.
150
The results of Mautz's survey should not be totally
unexpected since it is a "chicken or the egg" type of
problem. That is, appropriate data cannot be generated, or
even defined, before an appropriate theory of value is
defined; and an appropriate theory of value cannot be imple-
mented without appropriate data. Thus, unless a generally
accepted theory of value for valuing conglomerate companies
is available, then no survey can be conducted which will lead
to unequivocal conclusions.
Since no generally accepted theory of value has been
developed for valuing traditional types of companies, it is
doubtful that such a theory is imminent regarding conglome-
rates; however, it is hoped that the theory of value presented
earlier and implemented in the last chapter is at least
indicative of possible theories which could be adopted. If
nothing else, perhaps the nature of the problem will be
illuminated.
Data Required for Valuing Conglomerates According to the Value Added Criterion
Since one of the major sources of confusion regarding
conglomerates is their ability to obtain income from both
operating sources and through acquisitions, it follows that
a desired outcome from the accounting data would be a
reliable presentation of the total operating trend broken
down into acquisition and operating components. Ideally,
such information would enable investors to make reasonable
151
projections of future values for these two components of
income. However, since the past is the only guide to the
future, a reasonable presentation of historical data is the
best that can be expected. The historical data might be
broken down meaningfully into categories such as the
following:
(1) A statement of the company's acquisition strategy
in terms of relative values for Ep, Tes, Eg, gp, gs, trend
in size and rate of acquisitions along with supporting
financial information.
(2) A discussion of the sources of value added by the
parent company to the subsidiaries it has acquired, along
with a comparison of the rate of return on investment which
occurred after acquisition with that which would have occurred
if the subsidiaries and parent had not been combined.
(3) A statement of the estimated operating trend if no
acquisitions are made.
(4) A statement of the company's policy, if any,
regarding changes in accounting methods which affect the
computation of income and rate of return on investment
received from a subsidiary after it is acquired.
While the suggestions on the preceding page and above
regarding the information needs of investors are not
necessarily complete or final, such information would facili-
tate the making of informed projections of operating and
acquisition income. Unfortunately, the information is not
152
currently being provided, and the attainment of valid
inferences about such companies is made difficult because
of it. Obviously, the conclusions that can be drawn cannot
be more comprehensive and more valid than the data upon which
they are based.
Modifications to the Concept of "Value Added"
Imposed by the Forementioned Problem Areas
In the preceding chapter, negative value added was
defined as occurring anytime the "k" in the following
equation was equal to or less than one:
Rt = K (Term A' + Term B')
where
dp) (Ri„> Term A* =
a=t JP + 5 1 Ia(1+g]Ca)t"k
a~o
and
Ia(l+gIa)t~a'Ria(
1+gRIa)t_a
Term B' = ^ 5 • l a a
a=t
a=o
If "k" is less than one, then the collective contri-
bution to rate of return on investment by the following
sources is negative:
(1) From internal expansion of the parent company
(2) From divergences from the operating trend implied
by the subsidiaries prior to acquisition
153
(3) From synergy release.
To help clarify further the nature of "value added,"
and to implement the discussion in a more felicitious manner,
let attention be directed to the time periods described
below. The three time periods represent the growth periods
of subsidiaries prior to their acquisition, the period of
acquisition by a parent conglomerate, and an investment period
in shares of the conglomerate by an investor, respectively.
R-j- (rate of return on total investment)
Subsidiary Growth
1955 Period
Future > Range of f Possible
_J Trends
Period of Growth of
1961 Conglome-- 1967 rates
Investment Period
1977
Fig. 27—Relevant time periods in the determination of "value added."
If "k" is less than one during the 1961-1967 period, then
relative to the trend developed by the subsidiaries acquired
(which could extend back to the 1955-1961 period) and rela-
tive to the rate of return of the parent company at the
beginning of 1961, the net contribution to efficiency of
investment (measured in terms of rate of return on invest-
ment) of the three sources of income listed on the preceding
page is negative.
154
Ideally, all of the subsidiaries acquired by the parent
company during the 1961-1967 period would be considered in
determining whether or not value was added--i_.e. , whether or
not "k" was greater or less than one; however, since con-
glomerate companies acquire both public and nonpublic
subsidiaries and since information is available only for
publicly owned companies prior to their acquisition, it is
not possible or feasible to give recognition to all of the
companies acquired.
As an alternative to recognition of all the subsidiaries
acquired by a parent company, it is possible to state the
conclusions about "value added," relative to certain groups
of subsidiaries. In anticipation of this contingency, the
hypotheses presented earlier were stated in terms of "value
added" to the publicly owned subsidiaries only. That is, a
universe for testing the hypotheses was defined such that
only those conglomerates acquiring the highest proportion of
publicly owned subsidiaries would be included—publicly owned
prior to acquisition. Also, the conclusions about "value
added" are relative only to the publicly owned subsidiaries.
In the following equation,
Rt - k
ct—
<V(IV + Ia<1+9Ia)t"a Rj a+gRla)1^ ^ v a=o a
a=k IP + Ia(
1+gIa)t"a
a=o
if "k" is less than one, and if the subsidiaries considered
155
are limited only to those publicly owned prior to acquisition,
then the changes in attributable to the following sources
of income are negative:
(1) Synergy
(2) Divergences from projected prowth trends of the
publicly owned subsidiaries
(3) Internal expension of the parent company
(4) Internal expansion and acquisitions of nonpublic
subsidiaries.
In other words, from the publicly owned subsidiaries' point
of view, the influence of the above sources of income on rate
of return on investment has been negative if "value added" is
negative.
To place the idea of "value added" into a more meaning-
ful perspective, it is possible to conceive of all of the
publicly owned subsidiaries as a single unit with a single
spokesman. If the spokesman demanded that the conglomerate
company maintain an efficiency of investment (either through
its internal expansion or through its acquisitions) which was
greater than that which would have occurred if the public
subsidiaries and initial parent had remained unchanged, then
he is demanding that the parent company add value to the
publicly owned subsidiaries.
156
"Value Added" from Stockholders Point of View
One basic assumption underlying the hypotheses to be
tested is that "value added" is best measured relative to
changes in rate of return on investment before and after
acquisition. The propriety of rate of return on investment
as a measure of "value added" is not an issue—since it is
an assumption; however, one point demands clarification—
namely, whether the calculation of "value added" should be
based on a before-tax or after-tax rate of return.
If attention is directed to before-tax rate of return
and "k" is negative, then the net contribution of "value
added" of the following sources (from the publicly owned
subsidiaries' point of view) is negative:
(1) Synergy
(2) Divergences from projected growth trends of the
publicly owned subsidiaries
(3) Internal expansion of the parent company
(4) Internal expansion and acquisitions of nonpublicly
owned subsidiaries.
If attention is shifted from a before-tax to an after-tax
rate of return on investment, then the effects of changes in
tax rates, debt-equity ratios, and debt charges also are
influential in determining whether or not "value" is added.
That is, if fixed charges per unit of investment rise after
acquisition, then they will contribute in a negative manner
to "value added."
157
The following is an expression for rate of return on
investment after taxes (where R .x refers to return after
tax) :
(1-T) I (Rt) dt)-(Rd) (D)]
It
or
Rtx
(1-T) [(Rt)-(Rd)(|^)] = R t x
Also,
T = Tax Rate
R-l- = Rate of Return on Investment
It = Total Investment
Rd = Rate Paid to Debt
D = Total Debt
The preceding equations indicate that changes in T, R^, and
cause R-|-x to be different from Rj-. Thus, as previously It
noted, if attention is shifted to rate of return on invest-
ment after taxes, then changes in tax rates, rate of debt
charges, and debt-equity ratios are also factors influencing
the result. If these changes are referred to as leverage
changes, and if no value is added to its publicly owned sub-
sidiaries by a conglomerate company, the collective effect
on R t x of the following sources is negative:
(1) Synergy
(2) Divergences from projected growth trends of the
publicly owned subsidiaries
(3) Internal expansion of the parent company
158
(4) Internal expansion and acquisitions of nonpublicly
owned subsidiaries
(5) Leverage changes.
Since a sizeable portion of the data available provides
only income data after taxes, after-tax rate of return is
used in testing the hypotheses. Accordingly, calculations
of "value added" are influenced by changes in all five of
the forementioned variables.
Testing the Hypotheses
In this section of the paper, a selected group of con-
glomerate companies is examined. For each of the companies
included, the following sets of information are desired:
(1) Whether or not "value added" to the publicly owned
subsidiaries is positive or negative,
(2) The trend in k as size and diversification have
increased,
(3) Whether or not the Rja's of the publicly owned
subsidiaries have been consistently higher than the Rj- of
the period prior to the one in which they were acquired
(If this is true the implication is that acquisition income
may have been one of the major motives of the parent.),
(4) The acquisition strategy of the company in terms
of the ratios 51-3. and for the public subsidiaries, Rt gRt '
(5) The total trend in rate of return on investment
from both acquisition income and operating income as a
159
percent of investment (This trend is used to make future
projections of earnings per share, which, in turn, allows
values to be placed on the shares.),
If the hypotheses presented earlier are correct, then
"value added" to the publicly owned subsidiaries will tend
to be negative. Also, "value added" will tend to decline as
size and diversification have increased—i.-©-/ "k" will
become more and more negative. If acquisitions have been
the major source of income to conglomerate companies, then
the Ria's would be expected to be consistently higher than
the Rt's of the prior period, and an upward trend in Rj 1 s
would not be entirely unexpected. If hypothesis four is
correct, then the companies with relatively similar patterns
of "value added" will tend to have similar acquisition
strategies—where such strategies are defined in terms of
the ratios of RIa and . Finally, if hypothesis number Rit gRit
three is correct, then, on the basis of the same operating
trend and the same acquisition trend, and if an 8 percent
rate of return in demanded, the current market value of the
stocks of the conglomerate companies is too high.
Teledyne Company
The first company to be examined is Teledyne Company.
This company, like most of the other conglomerates, is
experiencing its greatest growth in the decade of the 1960's,
The company's total assets in 1962 were slightly over 10.5
160
million, but by 1966 its assets were over 170 million.2
Obviously, the company had experienced a considerable growth
in the interim period. Also, the company's rate of return
on investment after taxes climbed from 3.05 percent in 1962
to 7.06 percent in 1966. Of course, this growth trend can
be attributed to some combination of both operating and
acquisition income. Since expansion can be achieved more
rapidly through external avenues than through internal
avenues, the rapid growth in total assets of Teledyne is
perhaps indicative of the major source of its growth.
The publicly owned companies upon which the calculation
of "value added" is based are Kinetics Corporation, Hydra
Power Corporation, Micro Wave Electronics, Geotecbnical
Corporation, Servomechanisms, United Electro Dynamics, Dubrow
Electronics Industries, and Sprague Engineering. All of the
companies were acquired by the parent company in the period
1962-1966. Before examining the financial data of these
companies, some of the financial characteristics of Teledyne
are summarized in Table V on the following page.
Summary of the Performance of Teledyne
Since the R . for Teledyne in the period 1963-1966 was
greater than that which would have occurred if the initial
parent (in 1962) and the publicly owned acquisitions had
remained unchanged, value was added. The fact that value
2See Table V, p. 161.
161
TABLE V
FINANCIAL CHARACTERISTICS OF TELEDYNE INCORPORATED
Year
It
Investment
&t
Rate of
Return
P
Number of
Preferred
RP
Return to
Preferred
M
1 Minority
Interest
j
Rm
Return to
Minority
S
Shares C.
Stock
1966 170.4X106 7.06% 624,000 $3.50 none none 2.74 x 106
Change
1965 66.5* 106 5.64% 72,000 1.00 none none 1.84 x ' 106
Change
1964 47.4X106 5.36% 74,156 1.00 none none 1.05 x
106
Change
1963 23.9X106 5.35% 92,827 O O
none none .840* 106
Change
1962 10.8X106 3.05% N.A. N.A. none none . 646x
106 Change
1961
Change
1960 Source: Moody's Industrial Manuals
was added is reflected in the values for "k" which are pre-
sented in the following pages. If the viewpoint of all the
subsidiaries had been taken, instead of just the publicly
owned ones, it is possible that a different result could
have been attained.
162
TABLE VI
ACQUISITIONS OF PUBLICLY OWNED COMPANIES USED IN DETERMINING "VALUE ADDED"
Year
Kinetics Corp., acquired in 1966
Hydra Power Corp.,
acquired in 1965
Micro Wave Electronics,
acquired in 1965
Ja RIa la RIa la RIa
1965 1.38x106 13.56% • • • • • • • • • • • •
1964 1.38x106 8.32% 2.80x106 8.11% 2.61x106 10.79%
1963 1.27x106 7.91% 2.19x1C)6 7.35% 2.03x106 16.97%
1962 2.27x106 3.32% 2.08x106 5.48% 1.71*106 11.43%
1961 2.10xl()6 3.21% 2.08x106 -1.85% .701x106 25.97%
1961 » • • 2.00x106 4.39% N.A. N.A.
Geotechnical Corp.,
acquired in 1965
Servomechanisms, acquired in 1964
United Electro Dynamics
acquired in 1964
la RIa la RIa la RIa
1964 4.74x106 8.19% • • • • » • • • • • • •
1963 4.66x106 6.49% 3.52*106 -14.72% 7.28x106 8.76%
1962 4.85x106 9.41% 5.04x106 - 3.89% 7.63x106 3.85%
1961 3.73x106 9.64% 6.63x106 7.76% 6.89x106 5.48%
1960 N. A. N.A. 7.46x106 -12.99% 5.66x106 7.68%
1959 • • • 9.22x106 - 2.11% 2.77x106 5.09%
Source: Moody's Industrial Manuals
TABLE VI—Continued
163
McCormick Selph Dubrow Sprague Associates, Electronics Engineering,
Year acquired in 1964 Industries, acquired in 1963 acquired in 1964
*a RIa -"-a RIa la RIa
1963 1.27x106 10.45% 2.91x106 6.75% • • • • • •
1962 1.6 x106 -7.70% 1.63x106 4.74% 3.95x106 3.88%
1961 N.A. N.A. 1.14x106 7.46% 3.05*106 4.09%
1960 N. A. N.A. N.A. N.A. 4.00x106 7.92%
1959 N. A. N.A. N.A. N.A. 4.00x106 8.16%
1958 N.A. N.A. N.A. N.A. 3.16x106 9.73%
As size and diversification increased, the value for "k1
appeared to decline. A "least squares" trend line would
indicate a slight decline in the rate at which value is
added—i,.e. , the trend in "k"; accordingly, value added
would appear to have declined as size and diversification
increased.
It is interesting to note that the rate of return in
the year of acquisition for every company considered was
greater than the R . of the prior period, except for
Servomechanisms. The overwhelming implication of this is
that acquisitions are designed to make immediate contri-
butions to overall rate of return.
164
TABLE VII*
PROJECTIONS BASED ON TRENDS DEVELOPED PRIOR TO ACQUISITION
Company 1963 1964
Itx106 (Rt) dt> It
X106 (It)(Rik)
Initial Parent (1962)
10. 8x1C)6 .331x106 10.8X106 .331x106
Sprague Engineering 3.82x106 .0764^106 3.88x106 .0167x106
Dubrow Electronics Industries
• • • • • • 3.65x106 .204X106
United Electro Dynamics
• • • • • • 9.34x106 .673*106
Servo-mechanisms • f t • • • 2.23x106 -.256x106
Micro Wave Electronics • • 1 • • • • ft • • • •
Geotechnical Corporation • • • • • • • • # • • •
Hydra Power Corporation • m • » • • • • • • • •
Kinetics Corporation • * • • • • • • • • • •
Total 14.62 .407 29.90 1.48
*These data are derived from information contained in Tables V and VI.
165
TABLE VII—Continued
Company 1965 1966
Itx106 (Rt)(It> I t x i o 6 (I t)(
Rl k)
Initial Parent (1962) 10 .8X106 .331x106 10. 8X106 .331x106
Sprague Engineering 3 .94x106 -.0256x106 4. 00x106 -.10800
Dubrow Electronics Industries
4 .53x106 .236x106 5. 41x106 .265x106
United Electro Dynamics
10 .44x106 .793x106 11. 5x106 .9085x106
Servo-mechanisms .85x106 -.lllx106 • • • • • •
Micro Wave Electronics 3 .26*106 .205x106 3. 86^106 .0888*106
Geotechnical Corporation 5 .20x106 .345x106 5. 49x106 .325x106
Hydra Power Corporation 2 .78x106 .276x106 2. 95x106 .339x106
Kinetics Corporation • • • • • • 1. 01x106 .151x106
Total 41.80 2.05 46.17 2.30
The following values for K are based upon a"least
squares" projection of the trends developed by the companies
prior to their acquisition:
166
For 1963: R t = K(Term A1 + Term B') or 5.35 - K(2.8)
K = 1.9
For 1964: 5.36 = K (4.
K = 1.1
For 1965: 5.64 = K (4.
K = 1.2
For 1966: 7.06 = K (4 .
K = 1.7
The following are the equations for the "least square
trend lines for each of the companies acquired:
Kinetics Corporation: RT = 7.26 + 2.57t (t denotes a time)
Hydra Power Corporation: Ri a = 4.87 + 1.66t
Micro Wave Electronics: Ri a = 16.29 - 2.00t
Geotechnical Corporation: Ri a = 8.44 - .36t
Servomechanisms: Ria = 6.74 - 1.6t
United Electro Dynamics: Ri a = 6.17 + .35t
Dubrow Electronics: Ri a ~ 6.31 - .36t
Sprague Engineering: Rj a = 6.75 - 1.58t
From the preceding equations, it can be observed that
in five of the eight cases considered, the companies had
developed negative trends in return on investment prior to
the acquisition. If these negative trends are indicative of
the operating trend for the overall company, the company is
in a position of relying solely on acquisitions for its
167
growth. Whether this is the case or not cannot be determined
from the data available. Recall that if the operating trend
of overall company remains negative, earnings per share will
eventually decline.
Earnings per share, denoted by Y ., is expressed by the
following equation:
(Rtx) (It) - (Rp) (P) - (Rm) (M) Yt = 1
^ S
In this equation, R-|-x refers to after-tax rate of return on
investment. The following is a summary of some of the inde-
pendent variables in the equation for Teledyne and how they
changed during the time period under consideration.
TABLE VIII
RATES OF CHANGE IN REAL VARIABLES FOR TELEDYNE
Time Period Change in It Change in Rtx
Change in Rtx
Change in It
1962-63 13.lx106 2.30% .175x10"6
1963-64 23.5X106 .01% .0004*10-6
1964-65 19.1x106 .28% . 015x10-6
1965-66 103.8*106 1.42% .0137x10~6
From the preceding data it is apparent that 1^ is
growing rather dramatically, R t x is increasing moderately,
but growth rate as a percent of investment is declining.
168
That is, per unit of investment, the rate of return on
investment is declining. As noted previously, it becomes
harder and harder to increase R . as size increases if a con-
siderable part of growth is derived from acquisitions.
If R o x is considered to be the initial rate of return
on investment after taxes, then future values for R^x can be
defined in terms of change in It as
ARt
Rtx = t^o ^It ^t IQ)]
In the preceding equation, IQ denotes initial investment,
and It denotes the future value for total investment. Also,
denotes the rate of change of R-j- with respect to It- In 11
the next equation, It is expressed as a function of time:
It = tlo + "tiT (t-toU Where
IQ denotes initial investment
tQ denotes initial point in time
t denotes future points in time
A*t denotes rate of change if It with respect to time. At
If the equation for It is substituted into the one for
Rtx in the previous equation, then
A.It R t x = [Rto + ( A I t ) ( ^ ~ ) (t-tQ)
If P is assumed to maintain a constant percentage relation-
ship to I|-, then P as a function of investment is
169
Pn (r?) (IJ
Likewise for minority interest and common stock of parent:
M0 M = j~ dt)
o
and
S = So Io (It>
If the preceding equations for Rtx, P, M, and S are
incorporated into the expression for Y ., it becomes
R t o + ( a i t j ( t _ t o ) (Rp)(|2) - (Rm)("°)
xo I0 So Io
TABLE IX
VALUES FOR LEVERAGE VARIABLES FOR TELEDYNE
Year P RP Rm M S It It It
1962 N.A. 1.00 none none .060
1963 3.9X10~3 1.00 none none .036
1964 1.57x10~3 1.00 none none . 022
1965 1.08x10~3 • O
O
none none .027
1966 3.70x10~3 3.50 none none .016
170
If both numerator and denominator are multiplied by I0, then
the equation becomes
Yt =
A^tx Alt (R0) CI0) + (XlT") ^o) (t-tQ) - (Rp) (PQ) ~ (Rm) (MQ)
So
For Teledyne, if the variables with subscripts of "o" are
considered to be values for 1966 and the rates of change are
considered to be the average ones for the 1962-1966 period,
then the equation reduces to
Y t = 4.4 + 1.2 (t - t Q) - .8
Using this equation to project future values for Yt
TABLE X
PRESENT VALUE OF EARNINGS OF TELEDYNE UNDER CONDITIONS OF FUTURE CERTAINTY
Year Earnings per Share
Present Value at 8%
Year Earnings per Share
Present Value at 8%
1967 $4.80 $4.46 1972 $10.80 $6.80
1968 6.00 5.16 1973 12.00 6.96
1969 7.20 5.69 1974 13.20 7.13
1970 8.40 6.22 1975 14.40 7.20
1971 9.60 6.53 1976 15.60 7.17
Typically, Teledyne pays no cash dividends. Accord-
ingly, for purposes of valuation, it is assumed that no
dividends are paid and that the present value of the
171
common stock is derived from its sale after a ten-year
holding period.
In 1966, the price-earnings multiple of Teledyne was
around thirty-four.^ This rather high multiple was apparently
predicated on the assumption that rapid increases in earnings
per share would be sustained into the future. Since the
growth of the company will likely decline in the future as
acquisition income becomes less significant as a source of
earnings per share, it is quite possible that a lower
multiple of earnings would be more appropriate. Accordingly,
the present values of the shares of Teledyne are calculated
below at two different multiples—one multiple reflecting
the present value if the same multiple prevails in 1976, and
another reflecting an 8-percent-equivalent multiple.
TABLE XI
PRESENT VALUE OF SHARES OF TELEDYNE UNDER DIFFERENT ASSUMED MULTIPLES OF EARNINGS
Price-Earnings Price in Present Value Multiple 1976 at 8 Percent
12.5 $195.00 $90.28 1
34 536.40 248.17 1
^Moody' s Handbook of Common Stocks, 1968, Moody's Investor Service, New York, 196¥, p. 906.
172
If the price-earnings multiple is used to approximate
the price which prevailed in early 1967, then a result of
$163.20 is derived. (Note that earnings per share reported
by Teledyne in 1967 were $2.12. per share, reflecting a 2 for
1 stock split and a stock dividend in 1967.^) Accordingly,
the valuations on the preceding page correspond to the value
of slightly more than two shares in 1967. On the basis of
these values it would appear that the price of the shares in
1966 was low if a 2 . multiple is appropriate for 1976, and
high if a £5 multiple is assumed appropriate.
An indication of the acquisition strategy of Teledyne
can possible be derived by examining the values for the
ratios reproduced below for the 1963-1966 period. In these
ratios,
Rla—represents the "normalized" average rate of return on investment of all resources acquired externally during the designated year
Rl^—represents the overall rate of return experienced in the prior year by the conglomerate company
gRla—refers to the "normalized" average growth rates of rate of return on investment for all companies acquired in a particular year
9Rlt—refers to the overall growth trend of the con-glomerate during the period
The ratio of 9RIa can be used to estimate the effects of gRit
acquisitions on the overall operating trend of the total
company, and if the ratio of is greater than one, the Kit
^Ibid., p. 905,
173
implication may be that the company is relying on acqui-
sitions for a considerable part of its growth.
TABLE XII
OPERATING AND ACQUISITION RATIOS FOR TELEDYNE COMPANY
1963 1964 1965 1966
RIa 2.01% .44% 7.59% 14.97%
ri; 3.05% 5.35% 5.36% 5.64%
gRIa -1.58 -.60 .01 2.57 g^it .50 .50 .50 ".50"
The preceding ratios indicate that in two of the four
years considered, Teledyne acquired companies whose normalized
rates of return on investment were greater than the rate of
return on investment of the overall company in the preceding
year. Since the ratios for the two remaining years showed
an opposite result, no clear inference can be made.
The second set of ratios is perhaps indicative of the
effects of acquisitions on operating income. That is, in
three of the four years considered, the growth trend in rate
of return on investment was less than the normalized trend
of the overall company. Perhaps the implication of these
results is that Teledyne seeks to acquire companies whose
overall operating growth trend is less than its own. Recall
174
that such companies are the ones which are generally the
easiest to acquire with a favorable immediate effect on
earnings per share.
Litton Industries
Litton, like all of the other conglomerates, has ex-
perienced dramatic growth in total resources employed during
the 1960's. In I960, its total assets were roughly 119
million.5 By the end of 1967, its total assets were in
excess of 945 million.6 Accordingly, the company grew by
nearly 800 percent during the time period under consideration.
From 1960 to 1967, the rate of return on total investment
rose from 6.3 percent to 7.4 percent.
The companies used in determining whether or not value
was added include Hewitt-Robbins, Adler Electronics, Royal
McBee Corporation, Clifton Precision Products Company,
Fitchburg Paper Company, Stouffer Foods Corporation, Landis
Tool Company, and Jefferson Electric Company. These com-
panies were all acquired by Litton in the 1962-66 time
period. The following is a summary of "value added" in each
of the related years:
For 1963: K = 1.04
For 1964: K = 1.16
^Moody's Industrial Manual, Moody's Investor Service, New York, June, 1967, p. 3075.
6Moody's Industrial Manual, Moody's Investor Service, New York, June, 1968, p. 2651.
175
For 1965: K = 1.15
For 1966: K = 1.06
According to the preceding values for K, it appears that
the overall rate of return experienced was greater than that
which would have occurred if the initial parent and public
subsidiaries had remained unchanged. Again, as in the case
of Teledyne, it is possible that a different result would
have been attained if all of the company's acquisitions
instead of just the public ones had been considered.
While value was apparently added to the publicly owned
subsidiaries, it appears that after 1964 the rate at which
value was added began to decline. But, for the entire period
under consideration, there was no substantial downward trend
established for the rate at which value was added.
The acquisition policy of Litton, as indicated by the
public acquisitions examined, is not clearly designed to make
immediate increases in the overall rate of return. Only in
four of the nine companies examined was the rate of return
experienced by the newly acquired company greater than that
for the combined company in the previous year. This result,
however, does not imply that acquisitions were not conducted
to make immediate contributions to earnings per share of the
parent.
The following are the equations for the least square
trend lines for each of the companies acquired:
Fitchburg Paper Company: Rja = 4.7 + (.05)(t)
176
Stouffer Foods Corporation: Ria = 7.2 + (.91) (t)
Hewitt-Robbins Incorporated: Ria = 2.9 + (.16) (t)
Adler Electronics: RI a = 3.6 + (2.0) (t)
Royal McBee Corporation: Rla = 1.96 + (.45) (t)
Clifton Precision Products: Rja = 10.06 + (.61) (t)
Landis Tool Company: Ria = 14.1 + (1-5) (t)
Jefferson Electric Company: Ria = 3.3 + (.46) (t)
It would appear from the trend lines developed by the
companies prior to their acquisition that Litton has been
rather selective in its acquisitions. As compared to the
experience of Teledyne, Litton has apparently concentrated
on acquiring companies having positive growth trends. Of
course, if all acquisitions instead of just the preceding
ones had been considered, it is possible that a different
picture of Litton's acquisition policy may have emerged.
Assuming that the company's growth per unit of invest
ment continues at the same rate, and also assuming that the
leverage position of the company remains the same in the
future as it was in 1966, then the following is the related
estimate for future levels of earnings per share:
Yt = 2.62 + .19 (t-tQ) - .02
In 1966, the price per share of Litton's common stock
fluctuated between an approximate range of $56 per share and
$86 per share, while at the same time earnings per share
were $2.60,^ hence the company's stock was selling at
7ibid., p. 2649
177
TABLE XIII
PRESENT VALUE OF STOCK OF LITTON UNDER CONDITIONS OF FUTURE CERTAINTY
Year Earnings per Share Present Value at 8%
1967 $2.79 $2.58
1968 2.98 2.39
1969 3.17 2.35
1970 3.36 2.34
1971 3.55 2.28
1972 3.74 2.23
1973 3.93 2.16
1974 4.02 2.12
1975 4.21 2.01
1976 4.40 1.94
approximately twenty-five tiroes current earnings. Assuming
that this same multiple prevails in 1976, then the stock
would sell for $110 in that year. The present value of $110
discounted back at 8 percent in $47.30 .in 1966.
As in the case of Teledyne, depending upon the price-
earnings multiple which is assumed to exist in 1976, the
stock can be made to appear over-valued or under-valued, as
indicated in Table XIV.
178
TABLE XIV
PRESENT VALUE OF SHARES OF LITTON UNDER DIFFERENT ASSUMED MULTIPLES OF EARNINGS
Price-Earnings Price in Present Value Multiple 1976 at 8%
5 $22.00 $10.19 1
12 o
00 CN
LO 24.45
1
25 110.00 47.30 1
Summary of the Performance of Litton
Since the R - for Litton in the period 1963-1966 was
greater than that which would have occurred if the initial
parent (in 1962) and the publicly owned acquisitions had
remained unchanged, value was added. The fact that value
was added is reflected in the values for K which were all
greater than one. If the viewpoint of all the acquisitions
had been taken, instead of just the publicly owned ones, it
is possible that a different result could have been attained.
For the entire period (1962-1966) the value of K showed
no distinct trend either upward or downward as size and
diversification increased. However, after 1964 consecutive
values for K began to decline.
With regard to the acquisition strategy implied by the
types of publicly owned companies acquired, no clear attempt
179
by the parent company to raise its overall rate of return on
investment can be inferred. Approximately one-half of the
acquisitions had a favorable effect on R^ in the year of
acquisition, and approximately one-half had an unfavorable
effect. The following ratios perhaps hold a clue to the
acquisition strategy of Litton.
TABLE XV
OPERATING AND ACQUISITION RATIOS FOR LITTON
1963 1964 1965 1966
RIa 8.5% 3.7% 3.3% 10.7% RIt 6.1% 6.6% 7.0% 7.5%
gRia .88 .16 .45 .96 gRit .30 .30 .30 .30
From the preceding ratios, it appears likely that the
parent company attempts to acquire companies whose operating
growth trend will contribute to the overall trend of the
parent. This is evidenced by the fact that the ratio of gRIa
gRlt
is greater than one in every year except .1964. From the
data used in this analysis, it is impossible to tell whether
or not the parent company was able to acquire these com-
panies without diluting earnings per share. It seems likely,
however, that the parent company may have been able to use
its relatively high P/E multiple to make acquisitions which
180
contributed positively to both operating income and acqui-
sition income.
It is perhaps possible to state that the valuation of
the common stock of Litton Company was too high in 1966 if
an investor demanded an 8 percent rate of return. This
conclusion rests on the valuation of $47.30, derived earlier.
That is, if the same amounts of income are derived in the
future as in the past from both operating and acquisition
sources, it is probable that the current valuation is too
high.
Textron
Textron is generally considered to be one of the
"synergistic" conglomerates. Like both Teledyne and Litton,
this company has experienced a considerable growth in total
assets employed during the 1960's. In 1962, this company's
assets were valued at $308,646,505.® By 1967, reported book
value of assets was $669,575,000.^ During the period 1962
to 1967, the company's rate of return on total assets rose
from 4.8 percent to 9.2 percent.
The acquisitions upon which the calculation of "value
added" is based are Parkersburg Aetna Corporation, Jones &
Lamson Machine Company, Sheaffer Pen Company, Bostitch
8Ibid., p. 1389.
^Ibid.
181
Company, and Gorham Corporation. The following represents .
the calculation of "value added":
For 1963: K = 1.27
For 1964: K = 1.70
For 1965: K = 1.78
For 1966: K = 1.83
For 1967: K = 1.92
On the basis of the preceding values for K, it appears
that the overall rate of return experienced was greater than
that which would have occurred if the initial parent and
public subsidiaries had remained unchanged. The following
are the equations for the least square trend lines for each
of the companies acquired:
Gorharn Corporation Rj = 3.94 + .59 (t)
Bostitch Corporation Rja = 9.1 + .44 (t)
Sheaffer Pen Company Rj = 3.74 - .47 (t)
Jones & Lamson Company Rja = -1.05 - .50 (t)
Parkersburg-Astna Rja = 3.98 + .41 (t)
Based upon the preceding trends and the average growth
rate for the parent company, the following ratios can be
determined.
No clear inference regarding the acquisition strategy
of Textron can be made from the ratios found in Table XVI.
If the ratios are indicative of acquisition strategy, then it
may be possible to state that acquisitions are not clearly
designed to make immediate contributions to overall rate of
182
TABLE XVI
OPERATING AND ACQUISITION RATIOS FOR TEXTRON
1963 1964 1965 1966 1967
R la Rlt
gRia gRlt
6.11% 4. 80%
41 . 88
•4.55% 6.20%
• .50 . 88
N. A.
N. A.
6.14% 7.80%
-.01 . 88
6.33% 8.80%
.59
.88
return on investment. Of the companies considered so far,
Textron appears to be the one most likely to be relying on
synergy for part of its growth.
Based upon the same growth rate experienced during the
period 1962 to 1967--and assuming that the leverage variables
remain around the average for the 1961 to 1967 period, then
the following equation can be used to make estimates of
future earnings per share:
Yt = 2.33 + .25 (t-tQ) - .02
In 1967, after a two-for-one stock split, the price
earnings multiple of the common stock of Textron was around
nineteen to one. 10 Also, the price range of the stock during
1967 was from a high of $55 to a low of $38.25.-*--'- Assuming
that a fifteen-to-one multiple exists in 1977, then the
10 Ibid., p. 1396.
-'-llbid.
183
TABLE XVII
PRESENT VALUE OF STOCK OF TEXTRON UNDER CONDITIONS OF FUTURE CERTAINTY
Year Earnings per Share Present Value at 8%
1968 $2.56 $2.37
1969 2.81 2.41
1970 3.06 2.43
1971 3.31 2.43
1972 3.56 2.42
1973 3. 81 2.40
1974 4.06 2.36
1975 4.31 2.33
1976 4.56 2.28
1977 4.81 2.23
common stock will be selling for $72.15 if earnings in that
year are $4.81 per share. If there is no payout of earnings,
the present value of revenues derived from sale of the stock
in 1977 would be $33.12.
Summary of the Performance of Textron
On the basis of the preceding analysis, it seems pos-
sible that Textron has added value to the subsidiaries it
has acquired. To be certain, of course, all of the acqui-
sitions instead of only the publicly owned ones would have
184
to be considered. Also, based upon the data available, value
added would appear to have increased as size and diversifi-
cation increased. Finally, based on the ratios of and
S^Ia, no clear strategy can be inferred regarding Textron's
acquisition strategy. Regarding the value of shares of
Textron, it is possible that in 1967 the market price was
overly optimistic if an 8-percent rate of return was demanded.
Walter Kidde and Company
This company has experienced rather dramatic growth in
both rate of return on investment and total investment during
the 1963-1967 period. In 1963, total assets were $26,327,780,
and rate of return on investment after taxes was 1.04 per-
cent. -*-2 By 1967, total assets had risen to $253,125,969,^3
and rate of return on investment had risen to 7.00 percent.
In 1967, price-earnings multiple was around twenty-four to
one, perhaps reflecting the market's expectation that the
same rate of growth was sustainable in the future.
The companies used in the calculation of "value added"
are M & D Store Fixtures, Pathe Equipment Company, Globe
Security, Dura Corporation, Sargent Company, and Lighting
Corporation of America. The following represents the calcu-
lation of value added:
^Moody's Industrial Manual, Moody's Investor Service, New York, 1964, p. 2745.
-^Moody's Industrial Manual, Moody's Investor Service, New York, 1968, p. 2428.
l4Ibid.
185
For 1964: K = 1.00
For 1965: K = 1.09
For 1966: K = 1.08
For 1967: K = .96
On the basis of the preceding values for K, it would appear
that a slight amount of value, on the average, is being added
to its acquisitions by Walter Kidde. However, it is doubtful
that "value added" is the major source of the firm's growth.
The following are the equations for the least square
trend lines for each of the companies acquired:
M & D Store Fixtures Rja = 7.89 + .34 (t)
Pathe Equipment Company Rja = 15.57 + 1.18 (t)
Lighting Corporation Rja = 8.01 + 1.53 (t)
Sargent & Company Ria = 3.27 - .11 (t)
Dura Corporation Ria = 4.61 + 1.01 (t)
Globe Security Company Ria = 17.33 - .64 (t)
Based upon the preceding trends and the average growth
rate for the parent company, the ratios presented in Table
XVIII were determined. If these ratios are indicative of
the acquisitions strategy of Walter Kidde, then it would
appear that acquisitions are designed to make an immediate
contribution to the company's overall rate of return. The
relative growth rates in rate of return on investment would
indicate that perhaps the company is relying heavily on
acquisitions and, to a lesser degree, on the operating trends
of the companies it acquires.
186
TABLE XVIII
OPERATING AND ACQUISITION RATIOS FOR WALTER KIDDE
1964 1965 1966 1967
Ria N. A. 15.11% 11.20% 8.40% Rit 2.58% 4.01% 6.94%
gRla N.A. .76 .15 -.23 gRit 1.50 1.50 1.50
Based upon the same growth rate experienced during the
period 1962 to 1967, and assuming that the leverage variables
remain the same as in 1967, then the following equation can
be used to estimate future earnings per share:
Yt = 3.61 + .77 (t - tQ) - .41
It is questionable whether or not the company can main-
tain its current growth trend since its growth seems to be
heavily oriented toward acquisition; accordingly, the use of
the current earnings multiple of twenty-four-to-one may not
be justified in the future. If this multiple is applied to
projected earnings per share for 1977, then the price per
share in that period would approximate $250 per share. The
present value of $250 discounted back at 8 percent is $115.50,
If the value of the shares is assumed to be derived from the
sale of the stock after ten years, then, since the price per
share in 1967 was around $73, it is possible that the shares
are undervalued. The preceding valuation assumes that the
187
TABLE XIX
PRESENT VALUE OF THE COMMON STOCK OF WALTER KIDDE UNDER CONDITIONS OF FUTURE UNCERTAINTY
Year Earnings per Share Present Value at 8%
1968 $3.97 $3.68
1969 4.74 4.06
1970 5.51 4.37
1971 6.28 4.62
1972 7.05 4.80
1973 7.82 4.93
1974 8.59 5.01
1975 9.36 5.05
1976 10.13 5.06
1977 10.90 5.05
same P/E multiple will prevail in the future as in 1967, and
that the company can maintain the same average growth through
both operating and acquisition sources in the future as in
the past.
Summary of the Performance of Walter Kidde
On the basis of the preceding analysis it would appear
that the company is relying rather heavily on its acquisitions
for growth, with very little or no value added subsequently.
Also, based on the data available, it would appear that the
188
rate at which value was added began to decline after 1965.
Finally, the current market price per share (in 1967) was
low if the assumption was made that the company would con-
tinue to sustain its past growth trend.
Gulf and Western
Gulf and Western is a conglomerate which has shown its
most significant growth in the past several years. In 1963/
total assets employed were reported to be $48,111,810. By
1967, total assets were $749,437,236.-^ Thus, in the period
between 1963 and 1967, Gulf and Western multiplied the value
of its assets by roughly nineteen times. During this same
time period, the rate of return on total assets after taxes
rose from 5.47 percent to 6.16 percent.
The acquisitions upon which the calculation of "value
added" is based are New Jersey Zinc Company, Paramount
Pictures, South Puerto Rico Sugar, North and Judd Company,
Collyer Insulated Wire Company, Desilu Productions, and
Universal America Company. The following represents the
calculation of "value added":
1964: K = 1. 00
1965: K = 1. 09
1966: K = 1. 27
1967: K = 1. 01
15Moody's Industrial Manual, Moody's Investor Service, New YorF7~TFF3", p. 572":
l^Moody's Industrial Manual, Moody's Investor Service, New YorFT~rgW, p. 2733"
189
On the basis of the preceding values for K, it appears that
the overall rate of return experienced was greater than that
which would have occurred if the initial parent and public
subsidiaries had remained unchanged.
The following are the equations for the least square
trend lines for each of the companies acquired:
Universal American
Desilu Productions
Collyer Insulated Wire
North and Judd Company
South Puerto Rico Sugar
Paramount Pictures
New Jersey Zinc
RI; 4.56 + .87 (t)
RL A = 3.33 + .24 (t)
R t = 2.14 + 1.19 (t) 8.
RIa
Rla
RIa
RIa
3.92 + .55 (t)
5.87 - .82 (t)
3.45 + 1.17 (t)
2.23 + .83 (t)
Based upon the preceding trends and the average growth
rate for the parent company, the following ratios can be
determined:
TABLE XX
OPERATING AND ACQUISITION RATIOS FOR GULF AND WESTERN
1964 1965 1966 1967
RIa RIt
N. A. 4.72% 5.05%
5.19% 5.29%
6.20% 6.83%
SRIa gRlt
N. A. .82 .17
.18
.17 ' .71 .17
190
If the preceding ratios are indicative of Gulf and
Western's acquisition strategy, then it would appear that
the companies acquired, on the average, tend to have about
the same rate of return as the parent company in the pre-
ceding year. That is, the implication may be that the
company seeks to acquire in such a manner so as not to dilute
significantly its overall rate of return in the year of
acquisition. By acquiring companies whose growth trends are
greater than the average for all the overall company, Gulf
and Western has apparently been able to report a rising
operating growth trend.
Assuming that the company continues to increase its
rate of return on investment, per unit of investment, at the
same rate as in the immediate past (.i.e. , prior to 1967) and
assuming that the leverage variables remain the same as in
1967, then the following equation can be used to make esti-
mates of future earnings per share:
Yt = 4.47 + .23 (t-t0) - .74
The price-earnings multiple of Gulf and Western common
stock has averaged around fourteen to one in recent periods.
Assuming that this multiple is the appropriate one in 1977,
and earnings in that year are $6.03, then the present value
of revenues from the sale of stock would be $39.08 if an
8-percent rate of return is demanded.
191
TABLE XXI
PRESENT VALUE OF FUTURE DIVIDENDS OF GULF AND WESTERN UNDER CONDITIONS OF FUTURE UNCERTAINTY
Year Earnings per Share Present Value at 8%
1968 $3.96 $3.67
1969 4.19 3.59
1970 4.42 3.51
1971 4.65 3.42
1972 4.88 3.32
1973 5.11 3.22
1974 5.34 3.11
1975 5.57 3.01
1976 5.80 2 .90
1977 6.03 2.79
Summary of the Performance of Gulf and Western
On the basis of the preceding analysis, it would appear
that Gulf and Western has added value to its acquisitions.
However, after 1966, as size and diversification increased,
the rate at which value was added appears to have declined.
With regard to the company's acquisition policy, it seems
possible that acquisitions are designed to contribute to the
growth rate of the overall company without diluting signifi-
cantly the company's current rate of return on investment.
192
Finally, since the market price of the company's stock sold
in late 1967 for around $53 per share, the shares may be
slightly over-valued.
Automatic Sprinkler Corporation of America
This company, like all of the others considered pre-
viously, has experienced a considerable growth in total
assets during the 1960's. The company was incorporated in
1963, and in 1964 its assets were $68,363,893.By the end
of 1967, total assets had grown to $158,135,000.^ During
the same time interval, the net increase in rate of return
on investment was from 5.13 percent in 1964 to 5.81 percent
in 1967.
The companies upon which the calculation of value added
is based are Baifield Industries, Scott Industries, Safeway
Steel Products, and Interstate Engineering Company. The
following represents the calculation of value added:
1965: K =1.00
1966: K = .79
1967: K = .54
On the basis of the preceding values for K, it would appear
that negative value was added, on the average, by the Auto-
matic Sprinkler Corporation.
- ^ M o o d y ' s Industrial Manual, Moody's Investor Service, New York, 1965, p. 1387.
ISMoody's Industrial Mannual, Moody's Investor Service, New York, 1968,p. 952.
193
The following are the least square trend lines for each
of the companies acquired:
Baifield Industries
Scott Industries
Safeway Steel Products
Interstate Engineering
R I a = 7.85 + 3.61 (t)
Rja = 2.40 + 3.39 (t)
Rla = 6.16 + 1.09 (t)
Rja = 11.44 - .50 (t)
Based upon the preceding trends and the average growth
rate for the parent company, the following ratios can be
determined:
TABLE XXII
OPERATING AND ACQUISITION RATIOS FOR AUTOMATIC SPRINKLER
1965 1966 1967
RIa RIt
N. A. 9.43% 9.28%
11.34% 7.41%
3RIa g^it
N. A. 1.09 .009
2.16 .009
If the preceding ratios are indicative of the acqui-
sition strategy of Automatic Sprinkler, then it would appear
that the company has acquired companies which contribute to
the overall operating growth trend of the company, as well as
immediately because of acquisition. Based upon the same
growth rate experienced during the period 1964 to 1967, and
assuming that the leverage variables remain the same as in
194
1967/ then the following equation can be used to estimate
future earnings per share:
Yt = 1.73 + .0025 (t - tQ) - .297
TABLE XXIII
PRESENT VALUE OF THE COMMON STOCK OF AUTOMATIC SPRINKLER UNDER CONDITIONS OF FUTURE UNCERTAINTY
Year Earnings per Share Present Value at 8%
1968 $1.43 $1.33
1969 1.44 1.23
1970 1.44 1.14
1971 1.44 1.06
1972 1.45 .99
1973 1.45 .91
1974 1.45 .85
1975 1.45 .78
1976 1.46 .73
1977 1.46 .68
Assuming that the company shares are selling at approximately
the same price earnings multiple in 1977 as in 196 7, then
the price will be roughly $32 per share. The present value
of $32, discounted back at 8 percent is about $15. Accord-
ingly, if the present value is considered to be derived
from sale of the stock in 1977, then the total present
195
value per share would appear to be about $15. Since the
company's stock is selling (in 1967) at a price higher than
this amount, it is possible that the stock is over-valued.
Summary of the Performance of Automatic Sprinkler
On the basis of the preceding analysis, it would appear
that the company is adding negative value to its acquisitions,
Also, it would appear that acquisitions are designed to add
to both operating and acquisition income. Finally, it
appears possible that the company's shares are over-valued.
CHAPTER VIII
SUMMARY AND CONCLUSIONS REGARDING "VALUE ADDED"
BY A SELECTED GROUP OF CONGLOMERATES
It was suggested in Chapter I that traditional valuation
methods are not applicable to conglomerate companies for
several reasons. Among these reasons was the consolidation
of operating income and acquisition income in financial
reporting—resulting in investor's inability to clearly
discern which was the major source of growth. It was further
suggested, by an illustration, that companies such as con-
glomerates could give the impression of being growth
companies while, in reality, their actual productivity was
declining.
It is possible to define several sources of growth
which are available to companies which engage in extensive
acquisition activity. Among these sources are
(1) Size of acquisitions
(2) Profitability of acquisitions
(3) Rate of acquisitions
(4) Synergy release
(5) Internal expansion of income
(6) Changes in the leverage position
196
197
Accordingly, earnings per share can be expected to
change with a change in one or more of the preceding sources
of growth.
Current methods used to value shares of stock in con-
glomerate companies generally do not give explicit recog-
nition to all of tl>e possible sources of growth listed on
the preceding page. In fact, it is impossible to determine
which of the categories of growth is primarily responsible
for the dramatic performance of conglomerates. In an effort
to help clarify the major sources of growth, a selected group
of conglomerates were chosen, and their performance tested,
relative to four hypotheses. These hypotheses are reproduced
below.
Hypotheses
(1) Conglomerate companies listed on the New York and
American stock exchanges acquiring the highest proportion
of publicly owned subsidiaries in the period 1961-1967 have
added negative value to them—where positive or negative
value added is measured relative to the weighted average of
the rates of return on book values of the resources employed
by the subsidiary companies prior to their acquisition. 4
Accordingly, negative value added is defined as the extent
by which the projected rate of return on investment exceeds
the rate actually experienced.
198
(2) The value added by conglomerate companies during
the period 1961-1967 has tended to decrease from year to year
as size and diversification have increased—where diversifi-
cation is measured by the number of different industries in
which resources are employed, and size is measured in terms
of the book value of consolidated assets. (The decrease in
value added can be from negative values to more negative
values or from positive values to less positive values.)
(3) Based on a projection of past acquisition policies,
past weighted average growth trends of parent and subsidiary
companies, and the same percentage of value added by parent
company organizations, the current market prices of the
common stocks of conglomerate companies are too high if an
8 percent rate of return is demanded.
(4) If the conglomerate companies surveyed are grouped
into (1) those adding the most value (relative to the other
companies) and (2) those adding the least value, then the
companies in each group will tend to have similar acquisition
strategies. Acquisition strategies for this purpose are
defined in terms of the relative rates of return and growth
rates of subsidiaries and parent companies at the time of
acquisition.
In Chapter VII, the hypotheses were tested. Value
added was defined as occurring any time that K was greater
than one—where K was the dependent variable in the following
relationship:
199
R+- = K
a=t
(Ip) (Rijg) + £Ia(1+gIa)t~a'RIa(l+gRIa)t-a
a^t a=t IP+§b
Ia(l+gIa)t~a 1 + Ia(l+gla)t a
P a=o
The variables were defined in the preceding equation such
that any time K is less than one, the contribution of the
following sources of income to rate of return on investment
are negative:
(1) Synergy release
(2) Divergences from projected growth trends of the
publicly owned subsidiaries
(3) Internal expansion of the parent company
(4) Internal expansion and acquisitions of nonpublicly
owned subsidiaries
(5) Leverage changes.
Accordingly, if K assumes a value of less than one, then the
collective effect on rate of return on investment of the
preceding variables is negative, and the only source of
increases in rate of return on investment is through the
company's acquisitions. Value added was defined as occurring
anytime that K was greater than one.
Testing of Hypothesis One
The results of testing hypothesis number one were some-
what surprising. Of the companies surveyed, only two of
them indicated negative value added in any of the years
tested. Hence, the preliminary conclusion would be that for
200
the limited group of conglomerates examined, value was added.
Such a conclusion cannot be extended generally to all con-
glomerates for several reasons. First, it is possible that
the group selected is not indicative of all conglomerates.
Second, since only the acquisitions which were publicly
owned prior to their acquisition were considered, it is
possible that a different conclusion would have resulted if
nonpublic acquisitions had also been considered.
If, on the average, value added is positive, this means
that, collectively, one or more of the following sources had
a positive effect on rate of return on investment during the
testing period:
(1) Synergy release
(2) Divergences from projected growth trends of the
publicly owned subsidiaries
(3) Internal expansion of the parent company
(4) Internal expansion and acquisitions of nonpublicly
owned subsidiaries
(5) Leverage changes.
One major source of bias in testing for value added was
the use of zero expansion rate for the initial parent. That
is, it was assumed that glp and gl*ip were zero during the
testing period.
In light of the closeness of K to a value of one in many
of the cases considered, if it had been assumed that the
initial parent had a small positive growth trend in operating
201
growth, probably almost all of the calculations would have
shown a negative value for K. In fact, it would be possible
to determine the internal growth rates of the initial parent
which would be needed in order for value added to be negative.
This could be done by setting K equal to one in the following
equation and solving for sets of values of glp and gRip which
satisfied the equation.
a=t
Rt=K Ip(l+glp)
t-RIp(l+gRIp)t 5_Ia^1+5Ia) a 'RIa (
1 +g RI a)t _ a
I (l+gIp)t+> Iad+gla)^3 I (l+gl Ia 0-+gIa)
t-a
- p v a=b a-o
Using the equation above, the values for glp and gRj could
be determined which caused K to be equal to one. Accordingly,
any values less than these would cause K to be less than one
and value added would be negative.
Because "divergences from the trend developed prior
to acquisition" is one of the variables which determines
whether or not K is greater or less than one, several
additional sources of bias are possibly introduced into the
analysis. Specifically, it is possible that changes in
accounting valuation methods could cause reported earnings
and balance sheet valuations to be different before and
after acquisition; also, changes in external operating con-
ditions could cause the trends in operating performance to
be different.
It was not practical, or possible, to correct for the
deficiencies related to changes in accounting methods
202
because the corrections would require a knowledge of the
methods used for all of the operating units and not just the
publicly owned ones. That is, in the calculation of K, the
trend in operating performance of the publicly owned sub-
sidiaries which was developed prior to their acquisition is
compared to that of the overall company after acquisition.
In many cases, the companies acquired hundreds of smaller
companies for which the data is unavilable. Even if the
data were available, the time involved in making the cor-
rections would not insure that the resulting data would be
comparable. The data would not necessarily be comparable
because changes in external conditions as well as changes in
accounting methods can significantly influence the end result.
To the extent that the external conditions which prevailed
during the development of the operating trends of the sub-
sidiaries prior to their acquisition is not necessarily the
same as that which prevailed after acquisition, the validity
and consistency of the results is diminished.
In summary, consistency of accounting methods and
changes in external conditions represent two limitations
implicit in the definition of value added which cannot be
circumvented. Hence, it is possible that the positive values
for K reflect favorable changes in business conditions
during the 1961-1967 time period, or possibly, the positive
values for K reflect changes in accounting valuation methods
before and after acquisition, or finally, it is possible
203
that the favorable values for K reflect holding gains due
to rising price levels.
Testing Hypothesis Two
Since generally both size and diversification increased
with time for all of the conglomerate companies considered,
the testing of hypothesis two amounts largely to an exami-
nation of what happened to the value of K for each of the
companies during consecutive years of the testing period.
Considering the experience of all of the companies together,
the number of times that K increased from year to year was
roughly the same as the number of times it decreased.
Specifically, the ratio of increases to decreases was five
to four. That is, for every five years in which an increase
occurred in the rate at which value was added, there were
four years in which declines occurred. Hence, for the con-
glomerate companies considered, value added apparently does
not depend only upon size and degree of diversification.
Testing Hypothesis Three
The results of testing hypothesis number three are
summarized in Table XXIV. The present values shown are the
ones calculated in the preceding chapter. The other data
regarding prices and earnings were taken from the sources
indicated at the bottom of the table.
Based upon the data presented in Table XXIV and the
assumptions implicit in the model used to project earnings
TABLE XXIV
COMPARISON OF PRESENT VALUE AND PRICE IN THE YEAR OF VALUATION
204
Year Name of Company
Approximate Price Range per Share
Median Price
Present Value
1966 Teledyne $133-$50 $91.50 $248.00
1966 Litton 86- 56 71.00 47.30
1967 Textron 55- 25 40.00 33.12
1967 Walter Kidde
79- 44 61.50 115.50
1967 Gulf & Western
62- 29 45.50 39.08
1967
C i "i v /•—« /-» <
Automatic Sprinkler
75- 26 50.50 15.00
Moody's Handbook of Common Stocks Standard and Poor's Security Owner1s Stock Guide "
per share--as well as the assumption that the value of the
shares of conglomerate companies are derived from sale of
the stock after ten years—'the shares of stock were over-
valued in four of the six cases (where the median market
price in the year of valuation is compared to the corres-
ponding present value). Accordingly, depending upon the
validity of the assumptions used in deriving the present
values, it is possible that the common stocks of the companies
considered were over-valued at the time of valuation.
205
Testing Hypothesis Four
The purpose of this hypothesis was to determine if any
clear pattern existed in acquisition policy among the com-
panies considered, and to determine if any patterns existed
between value added and certain strategies. To begin with,
no clear pattern in acquisition strategy was observed for the
companies as a class (where acquisition strategies were
assumed to be measured by relative values for — a and S XsL) . Rlt gRit
The ratio RIa was greater than one exactly one-half of the Rlt
time, and, likewise, the ratio of 9RIa was greater than one gKl~t
exactly one-half of the time. Accordingly, for the companies
considered, it cannot be concluded that they show any
pattern, collectively, in their acquisition strategies.
Table XXV is useful for summarizing the results of
testing for individual relationships between the ratio of ~a. Rlt
and value added. Recall that one of the basic assumptions
underlying the statement of hypothesis four was that com-
panies with high values of would tend to have low RIt
amounts of value added because they were presumed to be
concentrating on acquisitions rather than on operating
efficiencies.
The results illustrated in Table XXV tend to support
hypothesis four. That is, the companies whose ratios for R la RI't
tended to be greater than one were also the companies which
had average values for K (value added) which were less than
those experienced by companies in the other two categories.
206
• TABLE XXV
VALUE ADDED AND ACQUISITION STRATEGIES
Companies Whose
Rlt was
ompanies Whose Fla pit
Greater Than One Most of the Time
Pendency to be Greater or Less
Showed No
K* rhan One K K
Teledyne 1.48 Textron 1.70
Walter Kidde L. 03 Litton 1.10 Gulf and 1.09 Western
Automatic .78 Sprinkler
Average K L. 03 Average K 1.12 Average K 1.39
Companies Whose Rla was Less Than Rlt
One Most of the Time
years examined,
Perhaps the implication of this result is that companies
with relatively high ratios of are concentrating more on RIt
obtaining acquisition income and less on adding value to
companies after their acquisition.
As shown in Table XXVI, those companies having ratios
of 9&la which were greater than one tended to have values g^it
for K (value added) which were less than those for the com-panies whose SBJLsl ratios were less than one. Perhaps the
gRlt
implication of the latter set of results is that companies
acquiring other companies with operating trends which are
less than their overall trends are in a better position to
add value to them than in other instances.
207
TABLE XXVI
VALUE ADDED AND ACQUISITION STRATEGIES
Companies Whose
was Greater gRlt Than One Most of the Time K*
Companies Whose
Showed No gRit Tendency to be Greater or Less Than One K
Companies Whose
|£l| was Less
Than One Most of the Time
K
Gulf and Automatic Western 1.09 Sprinkler .78 Teledyne 1.48
Litton 1.10 Walter Kidde 1.03
Textron 1.70
Average K 1.09 Average K . 78 Average K 1.40
*K represents t he
CHAPTER IX
VALUATION ACCORDING TO EXPECTED PRESENT VALUE
The purpose of this chapter is to outline and illustrate
the implementation of the expected present value model.
Recall that, for conglomerate companies, no long-run trends
can be established which include periods of adverse business.
That is, conglomerate companies are a relatively new
phenomenon, having had their principal growth since the
1960-1961 recession. During this period, there has been a
fairly constant external business environment in which sus-
tained growth and inflation have been the major characte-
ristics. A trend based upon the experience of conglomerates
during the 1960's would, like other trends, be based upon the
assumption that the future environment and business activi-
ties would, on the average, be a continuation of those which
characterized the past.
One of the assumptions underlying the adoption of the
expected present value model for valuing conglomerates is
that the experience of the 1960's is not a fair indication
of what can be expected in the future. Two major reasons
can be cited for the inadequacy of the 1960's as an index of
future performance. First, it is by no means a certainty
that growth and inflation of the same average dimensions will
208
209
continue to dominate the economy. Second, given the same
conditions in the future, it is doubtful that acquisition
income can continue to be as significant a part of income
since good merger candidates will become more and more
scarce, and, for a given firm, sheer size demanded will
eventually inhibit it.
For reasons such as those mentioned in the preceding
paragraph, valuations based on a simple projection of past
trends is not considered appropriate. As a replacement for
making valuations on the basis of simple past trends, the
expected present value model was introduced in an earlier
chapter. The expected present value model does not rely on
projections of the past, but, rather, on a simulation of
future conditions and the effects of different conditions on
earnings per share.
Instead of assuming that the growth trend developed in
the past will continue in the future, the expected present
value model allows for recognition of any number of trends
considered possible. Each of the trend lines given recog-
nition must be defined in terms of a particular slope,
which corresponds to a different state of the world simulated,
In addition to a trend line for each state of the world
considered possible, a probability (Prn) must be defined in
order to derive the related expected present value.
As noted previously, if an investor were capable of
deriving perfect expectations of the effects of different
210
conditions on earnings per share, then trend lines repre-
senting different states of the world could be defined with
accuracy. Also, in investor with perfect expectations would
be able to derive accurate probabilities for each trend line.
Of course, in reality, perfect expectations are extremely
improbable.
Valuations in Terms of Expectations
While much issue can be taken with theories of value
and decision models which demand that values for future
states of the world be projected, the alternatives to these
theories are even less desirable. That is, if it is assumed
that reasonable values for future states of the world cannot
be anticipated, then no intelligent decision is possible.
If one treats the independent variables in theories of
value as expectations, then, in as much as expectations are
real at the time of decision, it is possible that no issue
can be taken with resultant valuations. In other words, if
decisions and valuations are made in terms of expectations,
it is not illogical to define value in terms of expectations.
If this is done, a security would be considered to have value
at the time of a decision, not because of its actual income
potential, but because of the expectations of income assigned
to it. This point of view is a departure from traditional
analysis, which attempts to define value as something
intrinsic in the security and not as an attribute of the per-
son placing the value.
211
The traditional point of view seems to concentrate more
heavily on the mechanics and models of the valuation process
and less on relationships and assumptions underlying the
specific models. Perhaps, a more meaningful point of view
would be to treat valuation processes as formalized methods
for investors to place values on investments in terms of
their own objectives and expectations about the investments.
In other words, a valuation process can be thought of as a
process in which an individual
(1) defines and clarifies his investment objectives;
(2) establishes a relationship between his objectives, alternatives, and states of the world; and
(3) places values on alternative investments in terms of numerical values for objectives and in terms of expectations about states of the world.
While a valuation can be shown to be wrong after the
fact, an investor not having the benefit of hindsight cannot
expect to place values which are perfect in retrospect. If
an investor performs steps one and two listed above in a
reasonable manner, then the only limitations to his valuations
would be those attributable to three. Limiting errors to
imperfect expectations is probably adequate reason for con-
ducting formalized valuation procedures; however, if values
are defined as being derived from expectations, no issue can
be taken with them at all. Or stated differently, given
that the information, objectives, and assumptions of a
particular investor are accurately incorporated into a
212
theory of value, then the valuations which result are the
best ones possible.
Expected Present Value Model
The following equations can be used to define those
variables which influence earnings per share of conglomerate
companies; accordingly, they indicate which factors should
be the dependent variables in a simulation of the effects of
different states of the world on earnings per share.
(l-T)[(Rt) (It)-(Fd) (D3-(Rp) (P)-(Rm) (M) yt
When used for predicting future values for earnings per
share, the Rt and It in the preceding equation depends upon
current levels of operating income and on the size and growth
trends derived from both future operating income and future
acquisitions. If a subscript of "p" denotes both the parent
company and subsidiaries held at a particular point in time,
and "a" denotes average values for future acquisitions per
year, then (R^)(It) c a n be expressed as
t=n (Rt) (It) = I (l+gIp)t.R (l+gRlp)t +^I a(l+gI a)t- R l (l+gRla)t
F t=o
In this equation, the subscript "a" denotes average values
per year for n years of acquisitions.
If the leverage variables D, P, and M are expressed as
fractions of I^, then the following are the variables which
determine Yj-:
213
Ip/ ^Ip' 9- Ipf Ja' ^Ia' 5- -Ia
T^' Tt' 1^' R d' RP' Rrn' T' s' a n d fc
In the simulation of different states of the world on Yt,
the effects are translated through the variables listed
above. That is, anyone wishing to evaluate the effects of
changes in conditions must be able to evaluate the effects
of such changes on the variables above. Once this is done,
then Yt can be determined since it is the dependent variable.
For purposes of discussion let it be assumed that the
company being valued has resources employed in four distinct
industries, and, furthermore, let it be assumed that each of
these industries has a different sensitivity to changes in
external conditions. (Presumably, one of the advantages of
conglomerate companies over other forms is that they provide
a degree of diversification not normally provided.)
Letting a subscript of one, two, three, and four denote
values for each of the four industries, then, at any point
in time,
Ip = I p l + I p 2 + Ip3 + Ip4
and
Rip - (^Ipi) (1-pl) + (^Ip2^ (^p2) (^Ip3) (Ip3> + (RIp4) (Ip4^
Each of the industry groups denoted with subscripts of
one, two, three, or four in the preceding equation would
have its own individual growth rates in both rate of return
214
on investment and total investment; furthermore, these rates
would be expected to change with a given change in external
conditions. Accordingly, the effects of changes in external
circumstances on Y-j- would be channeled through their effects
on individual industries. Thus, if future expectations are
to recognize the uniqueness of each operating area in which
a conglomerate company has resources employed, then Ip, glp,
RIp' an(^ gRIp must be divided into various industry com-
ponents .
For purposes of simulation, the components of future
income are divided into
(1) Gross income from parent and prior acquisitions
(2) Gross income from future acquisitions
(3) Changes in the levels of R<j, Rp, and
r> p M S (4) Leverage changes reflected in T, — , and
It 1t It It
In matrix form, a symbolic representation of the simulation
model recognizing four possible states of the world might
appear similar to that on the following page. The format of
the matrix can be used to organize the steps to be followed
in developing the simulation model according to the following
pattern:
(1) Simulation of future levels of gross operating income to be derived from parent and prior acquisitions
(2) Simulation of future income to be derived from acquisitions
(3) Simulation of changes in the levels of R^, Rp, and Rm
215
Probability .50 17 .17 16
State of the World (1) (2) (3) (4)
Conditional Gross Income from Future Acquisitions
Conditional Gross Income from Parent and Prior Acquisitions
Industry: Industry: Industry: Industry: Conditional Gross Income from Parent and Prior Acquisitions
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Conditional Levels of Rd' Rp t Rm
Conditional Levels of T, D , P , M , and
It It It S It
Conditional Earnings per Share
Present Value of Conditional Earnings per Share (PVn)
n=4 EPV = 2 1 (prn) (pVn)
n=l
Fig. 28--Valuation Matrix under Conditions of Future Uncertainty.
216
(4) Simulation of leverage changes reflected in values for T, D , P , M , and S .
It It It It
Before actually attempting to simulate conditions
according to the format above, it is probably important to
point out certain problem areas and assumptions which must
be considered. First of all, a simulation of the components
of income is necessary because the values for the variables
which determine earnings per share are assumed to be gene-
rated under conditions of future uncertainty. Accordingly,
the individual placing the value must be able to specify his
expectations in terms of probabilities. Presumably, such
expectations would be based upon a consideration of all the
factors which influence the particular variable being
simulated—both internal and external to the company.
A second point worth noting is that the simulation
period is assumed to be ten years; accordingly, the growth
trends expressed in terms of probabilities will be average
ones for a fairly long period of time, which logically would
reflect expectations which were secular rather than cyclical
or seasonal in nature. After expectations have been defined,
then, in so far as the variables might be expected to move
in direct relationships to one another, the possibility of
"covariance" must be recognized.
217
Simulation of Future Operating Income from'
Parent and Prior Acquisitions
If it is assumed that the company being analyzed has
resources employed in four different industries whose growth
rates and profitability are different, then the effects of
changes in expectations must be specified for each company
or operating group within a particular industry. The fol-
lowing definitions are a necessary starting point in this
direction: Rpl—current rate of return on investment of resources
employed in industry number one
r 2—current rate of return on investment of resources employed in industry number two
R 3—current rate of return on investment of resources employed in industry number three
Rp4—current rate of return on investment of resources employed in industry number four
Ipl—current book value of resources employed in industry number one
Ip2—current book value of resources employed in industry number two
Ip3—current book value of resources employed in industry number three
Ip4—current book value of resources employed in industry number four
At the time valuation is to occur, it is assumed that
the rates of return on investment in each of the four oper-
ating areas are known, along with the initial total investment
in each of the operating areas. The unknown variables are
future growth rates for rate of return on investment and
218
for rate of investment in each of the four operating areas.
Obviously, such variables are subject to future uncertainty,
and related expectations must be expressed in terms of
probabilities.
Any number of approaches are possible for defining expec-
tations about future growth rates. For example, if there
are a large number of industries in which a particular
company is operating, then it might be advisable to define
overall expectations, such as for all nonfinancial corpo-
rations. Specifically, if it is assumed that in each of the
four industries under consideration, the rate of investment
moves in the same direction as changes in the overall rate,
but with 110 percent of the overall rate in industry number
one, in industry number two with 130 percent of the overall
rate, and in industries three and four with 80 percent and
90 percent of the overall rate, respectively. Then, the
following represents the generalized expectations for changes
within each industry.
Industry Number Expected Growth Rate of Internal Investment for each Industry as a Percent of Expectation for All Non-financial Corporations (glnf)
1 (1.10) (glnf)
2 (1.30) (glnf)
3 (0.80) (glnf)
4 (0.90) (glnf)
219
In a manner similar to that just described, relation-
ships between overall expectations and expectations within
single industries can be derived for rates of return on
investment. Accordingly,, if it is assumed that rate of
return on investment in each of the industries responds in
the same direction but with only 80 percent, 90 percent,
110 percent, and 70 percent of the magnitude, then changes
in growth rates of rate of return on investment can be
denoted for each industry as
Industry Number Expected Growth Rate of Rate of Return on Investment as a Percent of Expectation for All Nonfinancial Corporations (gRinf)
1 (0.80) (gRinf)
2 (0.90) (gRInf)
3 (1.10) (gRinf)
4 (0.70) (gRinf)
If the resources which the company has employed in the
four industries are expected to perform at the same .level as
industry averages, then the changes in the forementioned
variables for the industry averages can be used as the expec-
tations for the specific sets of resources. On the other
hand, if the individual operating groups are expected to do
better or worse than the industry averages, then allowances
must be made. For purposes of illustration, let it be
assumed that the resources employed in each of the four
220
industries are characterized by the following expectations
expressed as percentages of industry expectations.
Resource Group Number Expected Growth Rate of Internal Investment as a Percent of Expectation for Industry
1 80%
2 115%
3 75%
4 90%
Resource Group Number Expected Growth Rate of Rate of Return on Investment as a Per-cent of Expectation for Industry
1 50%
2 90%
3 110%
4 80%
Given the preceding assumptions, then the following are
expressions for expected growth rates of investment and rates
of return on investment for each of the four sets of resources
which the company employs:
gl p l = C-80)(1.10)glnf
glp2 = (1.15)(1.30)glnf
glp3 — (•75) (.80)glnf
glp4 (.90)(.90)glnf
gRIpl = (.50) (. 80)gRInf
9RIp2 = (-90)gRInf
gRjp3 — (1.10) (1.10)gRjnf
9RIp4 = (*80) (.70)gRjnf
221
In order to further develop the simulation model, let
it be assumed that the values for rate of return on invest-
ment and total investment for resources employed in each of
the industries are given by the following:-*-
RIpi 10% I p l $1,000,000
Rjp2 20% I p 2 $2,000,000
Rjp3 20% lp3 $1,500,000
Rlp4 12% Ip4 $4,000,000
Given the initial values listed above, the expectations
for future levels of operating income from each of the sets
of resources employed are presented below in general terms
Income From Industry Number
(1) = (.10) [1.01 +(.4)gRInf]10* (1x10s)[1.02+(S.8)gInf]
10
(2) = (.20) [.98+(.81)gRInf]10" (2xl06) [.97 + (1.495)glnf]
10
(3) = (.20) [.99+(1.21)gRInf]10' (1.5xl06)[1.02+(.60)glnfJ
1 0
(4) = (.12) I1.01+(.56)gRInf]10* (4xl06)[1.01+(.81)glnf]
10
Using these quations, a simulation of income patterns from
various industries can be obtained by simulating the expected
performance pattern for all nonfinancial corporations. The
reference distribution—in this case the performance of all
nonfinancial corporations—is something of an arbitrary
Note: The values above are values which exist at the time valuation is to occur and are assumed to be known by the individual attempting to make the valuation.
2Note: The simulation is for a ten-year period, which accounts for the exponents in the preceding expressions.
222
decision; however, this particular group was selected for
purposes of illustration because past data is available and
"covariance" relationships can be defined and supported more
appropriately for this reason.
Defining Expectations for Nonfinancial Corporations
Since the expectations of growth for the resources
employed by the company under consideration have been assumed
to be related to overall expectations for all nonfinancial
corporations, it follows that a simulation of possible growth
trends for nonfinancial corporations is equivalent to a simu-
lation of the growth trends for the groups of resources in
question. The next task to be dealt with is a simulation of
future possible growth patterns based upon the assumed
relationship between specific resources and overall expec-
tations for nonfinancial corporations.
As noted in an earlier chapter, the valuation of con-
glomerate companies is not a valuation in which the future
states of the world can be assumed to be known with
certainty. Accordingly, the recognition of future uncer-
tainty was deemed imperative. Furthermore, since the process
which is generating future states of the world seems to be
most appropriately described as an imperfect.one, a Bayesian
approach was considered appropriate. In line with this
reasoning, the person placing a value must be able to express
expectations in terms of probabilities. While any number of
223
approaches are possible, the following one is recommended
and is used for purposes of illustration.
First, since Bayesian decision theory relies on sub-
jective expectations, the initial requirement for valuing
shares in conglomerate companies is that all of the expec-
tations for the variables influencing be expressed in
terms of probability distributions. The first step in
defining such expectations is a specification of the following
for nonfinancial corporations:
(1) The most likely growth rates for rate of return on
investment and for rate of investment
(2) The most optimistic expectation for growth rates
(3) The least optimistic expectation for growth rates
In graphic form, the three expectations might appear similar
to the following:
Most Optimistic
Most Likely
Least Optimistic
t=0 1 2 3 4 5 6 7 8 9 10
Years
Fig. 29--Initial range of expectations
Each of the three trend lines presented above corre-
sponds to a particular growth rate. The lines presented
reflect linear growth trends; however, if it were felt to be
224
appropriate, the trend lines could be drawn to reflect
geometric growth trends. Also, the three expectations—
most likely, most optimistic, and least optimistic--would be
expressed for both rate of return on investment and for total
investment by all nonfinancial corporations.
The range between the most likely and the least likely
expectations contains all expectations of growth which are
considered as relevant possibilities; accordingly, the
summation of the subjective probabilities for growth rates
within this overall interval is one. After defining the
range of possible expectations, the second recommended step
is to bisect the angle between the most likely expectation
and the most optimistic expectation and, also, to bisect the
angle between the most likely expectation and the most
pessimistic expectation. This second step is reflected in
the following diagram.
^Range number 4
Range number 3
Range number 2
Range number 1 t=0 1 2 3 4 5 6 7 8 9 10
Years
Fig. 30—Bisection of the initial range of expectations
225
By continuing to bisect angles in the manner just de-
scribed, more and more intervals of expectations can be
defined. For purposes of illustration only one bisection is
performed and only four ranges of expectations are created.
The class boundaries and class medians of these four inter-
vals can be derived and expressed as different growth rates
reflecting the slope of the line which defines the boundary.
For example, if the most optimistic expectation for growth
of rate of return on investment for all nonfinancial corpo-
rations is 10 percent, the least optimistic is -5 percent,
and the most likely expectation is 7 percent; then, the
following class boundaries and class medians are defined.
Range Range Range Range
Class Median
Class Boundary
(1) (2) (3) (4)
-2% 4% 7.75% 9.25%
-5% L
1% 7% J
8.5% JL
10% J
Fig. 31--Range of Expectations for Rate of Return on Investment for All Nonfinancial Corporations.
The next step which must be performed is a specification
of the subjective probabilities related to the expectations
contained within each class. For example, if the subjective
probabilities for ranges 1, 2, 3, and 4 are estimated to be
.15, .15, .50, and .20, respectively, then the following
frequency polygon can be constructed:
226
Relative Frequency
. 50
.40
.30
. 20
.10
0 % 8. % 8. 5% 5% 1 % 7 % 8. 5%
Fig. 32—Expected growth rate in rate of return on investment (gRIng).
The preceding frequency polygon is for expected growth
in rate of return on investment for all nonfinancial corpo-
rations. In a manner identical to that which was used in
developing the frequency polygon above, a similar one for
expectations of growth rates in total investment can be
developed. The final product is presented below.
Relative Frequency
.60
.50
.40
.30
.20
.10
-3% 2% l
6% 9% i
12%
Fig. 33—Expected growth of investment (glnf)
227
The relative frequencies (i.e., subjective probabilities)
of the intervals from left to right in the preceding diagram,
are .20, .30, .30, and .20. Once frequency polygons such as
the preceding ones have been defined, then it is possible to
perform a simulation of the first component of future income
for the conglomerate. By allowing the class mean to be
representative of all of the expectations within a single
class, the simulation procedure can be simplified. If this
is done, then the implicit assumption is made that in the
long-run the expectations to the left and right of the
individual class means have equal probabilities.
To perform the simulation adequately, a computer program
is almost imperative; however, for purposes of illustration
a mechanical simulation under simplified assumptions is made.
To begin with, it was assumed that only four states of the
world were considered as relevant possibilities. Each of
these states would correspond to specific values for the
expectations contained in the preceding frequency distri-
butions. For purposes of illustration let it be assumed that
the four states of the world being simulated are charac-
terized by the following expectations:
(1) 9Rlnf 7.75% and glnf of 4%
(2) 9RInf 4% and glnf of 4%
(3) 9RInf 4% and glnf of 7.5%
(4) 9RInf o f 7.75% and glnf of 7.5%
228
By substituting the values into the equations on page
221 for glnf and gRinf which correspond to the four states
of the world being simulated, the income to be derived from
resources employed by the parent company at the time of
valuation are determined. Since the example is hypothetical
anyway, the numerical calculations are not carried out.
An Alternative Approach
Instead of relating expectations for the individual
groups of resources to the expectations for all nonfinancial
corporations a different, and perhaps more appropriate,
method can be conceived. Namely, it would be possible for
the analyst or other person performing the valuation to
express his expectations for each of the specific sets of
resources employed by the parent at the time of valuation.
These expectations could be defined by projecting again the
most likely, most optimistic, and least optimistic expec-
tations. However, in this case, these projections would be
for the individual sets of resources and not for all non-
financial corporations. Once such a projection was made,
frequency polygons could be created by bisecting the angles,
as before. In a manner such as this, probabilistic expec-
tations could be defined for growth in rate of return on
investment and growth in total investment for all sets of
resources employed at the time of valuation.
229
If individual frequency distributions are created for
specific operating units within the company, then special
allowance must be made for "covar.iance." To do this, con-
ditional expectations would have to be defined. For example,
if a particular conglomerate company operated only in two
industries, then the following steps would be followed in
defining probabilistic expectations:
(1) Select the industry whose expected growth of
investment and growth rate of return on investment is the
more stable, and use these expectations as the reference
distributions.
(2) Form the reference distributions by defining the
most likely trend, the least optimistic, and the most
optimistic trends. (Bisect the angles to create the desired
number of ranges.) Also, attach subjective probabilities to
the ranges created.
(3) Allow the class mean to represent each class
created in step two.
(4) For the second industry in which the firm is
operating, define conditional expectations based on given
values for growth rates corresponding to each of the class
means in the reference distribution. That is, for each of
the class means in the reference distribution, define the
conditional expectations for the second industry.
With only two industries, there will be two distri-
butions representing expectations for growth in rate of
230
return on investment and growth in total investment in the
reference industry. If the reference distributions contain
four classes each, then the conditional expectations for
industry number two can contain as many as eight distri-
butions—four reflecting conditional expectations for growth
in rate of return on investment and four reflecting con-
ditional expectations for growth in total investment.
By adopting this alternative procedure, any possible
covariance of growth rates is incorporated into the con-
ditional probability distributions. The only disadvantage
of such an approach is that the number of distributions
required increases geometrically as the number of industries
increases arithmetically. Note, that if specific sets of
resources operating within a particular industry have dif-
fering expectations, then separate expectations can be defined
just as if they were operating in separate industries.
Simulation of Acquisition Income
In order to simulate future levels of acquisition
income, or more accurately, income from future acquisitions,
the following variables must be endowed with probabilistic
expectations:
(1) I a—The average annual size of future acquisitions
(2) R I a—The average rate of return on investments in the year of acquisition for future acqui-sitions
(3) gla—Average annual rate of internal expansion of future acquisitions after acquisition
231
(4) gRja--Average annual growth in rate of return on investment after acquisition for future acquisitions.
The approach recommended for simulating the preceding
variables is identical in nature to that recommended as an
alternative method for simulating future income from re-
sources controlled by the company at the time of valuation.
Specifically, it is recommended that subjective probability
distributions be developed for each of the four variables,
based upon must likely expectations, most optimistic expec-
tations, and least optimistic expectations. Of course, as
before, the bisection of the angles between the various
expectations will lead to specific classes of expectations
for which subjective probabilities can be defined.
If it is considered desirable, covariance between the
four types of variables can be recognized by defining one of
them as the lead variable and developing conditional expec-
tations for the others. For purposes of illustration it will
be assumed that covariance between the four variables is not
considered relevant. Accordingly, the distributions pre-
sented in the following figures represent the end products
of the process described previously.
If the possibility of covariance is considered to be
significant, then these distributions can be used to simulate
future income streams to be derived from future acquisitions.
Again, to make an adequate simulation, the use of a computer
would probably be necessary; however, for purposes of
232
Relative Frequency
.60
.45
.30
.15
.00 .05
L 07 ...J
15 .25 1
. 35 I
Fig. 34—Range of possible expectations for Ia in millions of dollars.
Relative Frequency
. 60
.45
.30
.15
00 •5% i
-1% 1
3 % l
5% I
Fig. 35—Range of possible expectations for growth in total investment of future acquisitions after acquisition (gia>•
Relative Frequency
.60
.45
.30
.15
00 •2% i
3% i
8%m i
10% _J 12%
Fig. 36—Range of possible expectations for rate of growth in return on investment of acquisitions after acqui-sition (gRia)•
233
Relative Frequency
.60
.45
.30
.15
.00 2% U>
4% 6% _J 9 % I
12% J
Fig. 37—Range of possible expectations for rate of return on investment of acquisitions in the year of acqui-sition (Ria)•
illustration, it is assumed that only four states of the
world are considered possible These states are
la Rl a 9RIa
(1) .065x10® 4% 7. 5% 9%
(2) . 110x10® 6% 5 % 11%
(3) .200xl06 1% 3 % 5.5%
(4) .110x10s 6% 3 % 9%
By substituting the numerical values above into the
equation on page 212, the corresponding values for income
from acquisitions can be determined. The values for income
thus determined along with operating income previously
determined form the total gross income available to the con-
glomerate .
As noted previously, the variables which determine the
value of earnings per share for conglomerates are the fol-
lowing :
234
(1) Operating income from parent and subsidiaries
acquired prior to the point in time at which valuation is to
occur. This component of income is determined by Ip, Rip,
glp, and FRjp. Depending upon how many different markets a
firm is operating in, the preceding variables are used
several times to describe the operations of a single con-
glomerate .
(2) Acquisition income, or more specifically, income
from resources acquired externally in the future. This
component of total income in the future depends upon the
values for Ia, Ria, gla/ a n^ 9RIa*
(3) Rates of return paid to debt, preferred stock, and
holders of minority interest. These variables are denoted
by R(j, Rp, and Rrft, respectively.
(4) Leverage variables reflecting level of taxes, level
of debt, level of preferred stock, and level of minority
interest. These variables can be denoted by T,__D, P__, and It It
M , respectively. It
Since values for all of the preceding variables can be
viewed as being generated under conditions of future uncer-
tainty, related expectations must be stated in terms of
probabilities. So far, the variables included in one and
two above have been expressed in this manner, and a framework
for simulating different possible conditions has been es-
tablished. The next two sections of the paper are concerned
with three and four.
235
Simulation of 5kL' 5m
It would.be necessary for the person placing values on
the shares of conglomerate companies to define covariance
patterns and future expectations, or sets of expectations
for R^. The expectations for R<j could be short-run or long-
run in nature, depending upon the preferences of the indi-
vidual placing the values. In any case, it would be necessary
to calculate the average annual rate of interest paid for
each year of the ten-year simulation period and then calcu-
late the overall weighted average. This overall weighted
average would then be multiplied by total debt for each year
of the simulation to derive total annual interest payments.
This task (.i.e. / finding annual payments made to debt
holders) is no problem once the appropriate numerical values
for R^ and D have been derived; however, deriving the
numerical values may be a slightly more complex problem.
The approach which is recommended for simulating future
values for R^ is to relate the expectations for gRdnf (which
is hereby defined as the average annual percentage growth
rate of interest paid by nonfinancial corporations during
the ten-year valuation period) to the subjective expectations
previously defined for glnf- This recognition of covariance
is based upon the assumption that periods of rapid expansion
(characterized by relatively high values for glnf) will also
be periods in which debt financing and interest rates are at
relative highs. If this assumption is proper, then
236
conditional subjective expectations which recognize covariance
patterns between glnf an<3 gRdnf a r e n o^ unrealistic.
Accordingly, it would be necessary for conditional expec-
tations to be defined for gRdnf which would be predicated on
what values are selected for glnf. Following the same pro-
cedure used before to establish conditional expectations—
four probability distributions could be defined, each one
corresponding to the mean of the classes in the distribution
for glnf- Then, depending upon the value selected for glnf/
the corresponding value for gRdnf could be derived by
sampling from the related distribution.
One additional expectation would be required in order
to simulate values for R^; namely, it would be necessary to
relate expectations for changes in the growth rate paid on
debt by the particular company to expectations for all non-
financial corporations. For example, it might be assumed
that the particular company being valued always paid a
slightly lower rate of interest than the average rate paid
by all nonfinancial companies—in which case, a relationship
such as the following could be defined:
gRd = (.85)(gRdnf)
The relationship expressed in the preceding equation is that
the rate of interest paid is always 85 percent of the rate
paid by all nonfinancial corporations. Thus, to summarize,
Rd would be dependent upon the following:
237
(1) the value selected for g*nf/
(2) the value selected from the related probability distribution for gR^j, and
(3) the assumed relationship between gR<3n£ and gR^.
In the simulation of Rp and Rm several alternatives are
possible. It may be feasible to assume that the rate paid
to preferred, per share, remains the same over time. Or, if
it is considered desirable, subjective expectations can be
defined and used in simulating future levels for Rp. Also,
conditional expectations can be defined to reflect any
covariance which is believed to exist between Rp and other
variables influencing earnings per share. The same con-
ditions apply equally to Rm. For purposes of illustration it
is assumed that covariance is not significant and that
subjective expectations for both Rm and Rp are determined by
defining the most likely, most optimistic, and least opti-
mistic expectations. The following distributions are examples
of how the end result of the process might appear.
Relative Frequency
.60
.45
.30
.15
.00
Fig. 38--Yields on preferred stock (expected)
238
Relative Frequency
.60
.45
.30
.15
.00 I -$5 $0
™ 4 L , $5 $7 $9
_i Fig. 39—Rate paid per share to minority interest
(expected).
For purposes of illustration, it is assumed that the
four states of the world being simulated are characterized
by the following values for Rp and Rm:
Rr
(1)
( 2 )
(3)
(4)
$2.50
$3.50
$5.50
$3.50
R-•m
$6.00
$2.50
$8.00
$6.00
Since the change in the level of interest rates was
related to glnf/ it follows that states of the world 1, 2, 3,
and 4 will be dependent upon the related value for glnf«
Hence,
glnf Conditional expectations Related for gRdnf value—gRd
(1)
( 2 )
(3)
(4)
4%
4%
7.5%
7.5%
3.6%
3.6%
6.75%
6.75%
3.06%
3.06%
5.74%
5.74%
239
Simulation of-Leverage Variables
The remaining variables which must be simulated in order
for conditional values for earnings per share to be derived
are T, 5_, —., and L_. Subjective distributions for these It It It
variables including conditional expectations could be defined,
just as before. However, for purposes of illustration, it
is assumed that these variables maintain constant and known
values during the period of simulation. That is, regardless
of the state of the world, the following are the values which
are assumed to exist:
T—50%
5z~30%
It
P — 3 . 2 shares per $1000 of investment It
M — .6 shares per $1000 of investment It
S — 1 . 5 shares per $1000 of investment
It
With the addition of the preceding values for the
leverage variables, it is possible to fill in numerical values
for all of the desired components of the matrix on page 215.
Once conditional values have been determined for average
earnings per share for each of the states of the world
assumed to exist, then the average Y-t for each state of the
world can be weighted by its respective probability to
determine the expected present value. Since the example
under consideration is hypothetical and many iterations
240
would be necessary for the example to be meaningful, the
numerical calculations are not carried out.
In conclusion, it is probably worth noting that an
adequate simulation model along the lines just described
would require a computer. This is true because of the in-
ability to handle, mechanically, all of the variables and the
different possibilities which they represent. Instead of
including a computer simulation as part of this project,
consideration has been limited to a description of the
rudiments of the model which would be necessary to perform an
adequate simulation. A program for valuing shares in con-
glomerate companies along the lines just described has been
written and is currently in the final stages of refinement;
however, the program and its application to specific companies
is deferred and not considered to be a necessary part of this
project.
BIBLIOGRAPHY
Books
Bierman, Harold and others, Quantitative Analysis for Business Decisions, Homewood, Illinois, Richard D. Irwin, Inc., 1965.
Clendenin, John C. and George Christy, Introduction to Investments, 5th edition, New York, McGraw-Hill Inc., 1969.
Graham, Benjamin and others, Security Analysis, New York, McGraw-Hill Book Co., 1962.
Howell, James E. and Daniel Teichroew, Mathematical Analysis for Business Decisions, Homewood, Illinois, Richard D. Irwin, Inc., 1963.
Keynes, John M., The General Theory, New York, McGraw-Hill Book Co., 1964.
Madden, Edward H., The Structure of Scientific Thought, Boston, Houghton Mifflin Co., 1960.
Prather, Charles L., Financing Business Firms, Homewood, Illinois, Richard D. Irwin, Inc., 1961."
Schlaifer, Robert, Introduction to Statistics for Business Decisions, New York, McGraw-Hill Book Co., 1961.
Van Home, James C., Financial Management and Policy, Engle-wood Cliffs, New Jersey, Prentice-Hall Co., 1968.
Walker, Ernest W., Essentials of Financial Management, Englewood Cliffs, New Jersey, Prentice-Hali, Inc., 1965.
Williams, John Burr, Theory of Investment Value, Cambridge, Massachusetts, Harvard University Press, 1938.
Articles
Ackerman, Robert W. and Lionel L. Fray, "When the Return-on-Investment Criteria Are Applied to the Discounted Future Earnings of a Proposed Acquisition a Somewhat Truer Picture of Financial Value Emerges," Financial Executive, XXXV (October, 1967), 35-54.
241
242
Bjorksten, Johan, "Merger Lemons," Mergers and Acquisitions, I (Fall, 1965), 36-38.
"A Closer Look at Conglomerates," Financial World, CXXIX (February 14, 1968).
Briloff, Abraham J., "Distortions Arising from Pooling-of-interests Accounting," Financial Analysts Journal, XXIV (March-April, 1968), 71-80.
"Business Combinations," Accounting Research and Terminology Bulletins (Bulletin #48) , New York, American Institute of Certified Public Accountants, 1961, p. 22.
Christy, George A., "Does Today's Investor Appreciate the Diversified Firm's Significance," The Commercial and Financial Chronicle, CCIII (April 7, 1966), 1-20.
"Conglomerates," Forbes, CI (January 1, 1968), 53-55.
"Glamour Stocks that Ran into Trouble," U.S. News and WorId Report, LXIV (March 18, 1968), 112-113.
"How Much Data Must Conglomerates Bare?" Business Week, No. 2001 (January 20, 1968), pp. 53-55.
"Justice Agency Issues Merger Guide," The Wall Street Journal, Southwest edition, CLXXI (May 31, 1968), 1, 4.
Kitching, John, "Why Do Mergers Miscarry?" Harvard Business Review, XLV (November, December, 1967), 84-101.
Mautz, Robert K., "Findings and Recommendations Released on Financial Executives Research Foundation Study," Financial Executive, XXXVI (February, 1968), 18-28.
, "Identification of the Conglomerate Company," Financial Executive, XXXV (July, .1967), 53-64.
"Measuring the Surge of Mergers," Business Week, No. 2013 (March 23, 1968), p. 69.
"Mergers on Parade," Mergers and Acquisitions, III (March-April, 1968), 78-79. .
Patterson, James M. and D. Jeanne, "Conglomerates: The Legal Issues—A Review of Public Policy Needed," Business Horizons, XI (February, 1968), 39-48.
243
"Random Walks in Stock Market Prices," Financial Analysts Journal, XXI (September, October, 1965), 55-59.
Stabler, Charles N., "The Conglomerates Impact of Big Com-bines on the Economy Defies Oil-fashioned Analysis," The Wall Street Journal, Southwest edition, CLXVII (August 12, 1968), 1.
Tanner, James C., "Ling's Empire," The Wall Street Journal, Southwest edition, CLXX (August 18, 1967X7" !•
Wall, John, "Want to Get Rich Quick?" Barron's, XLVIII (February 5, 1968), 5-19.
Wendt, Paul F., "Current Growth Stock Valuation Methods," Financial Analysts Journal, XXI (March-April, 1965), 91-102.
Reports
Moody's Industrial Manuals, Moody's Investor Service, New York, 1957-1967.
Moody's Handbook of Common Stocks, Moody's Investor Service, New York, 1st, 2nd, and 3rd Quarterly Editions.
"Some Problems of Disclosure," excerpted from an address by Manuel Cohen, Chairman, Securities and Exchange Commission, March 7, 1968.
Standard and Poor's, Security Owner's Stock Guide, Standard and Poor's Corporation, New York, January, 1968-January, 1969.
10-K Reports to the Securities and Exchange Commission, Washington, D.C., 1964-1967: Teledyne Company, Textron Inc., and Automatic Sprinkler.